Professional Documents
Culture Documents
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or
by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no
expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of information
contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in
rendering legal, medical or any other professional services.
COMPUTATIONAL MATHEMATICS
AND ANALYSIS
FREDERICK J. MITCHELL
EDITOR
Copyright © 2017 by Nova Science Publishers, Inc.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted
in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying,
recording or otherwise without the written permission of the Publisher.
We have partnered with Copyright Clearance Center to make it easy for you to obtain permissions to
reuse content from this publication. Simply navigate to this publication’s page on Nova’s website and
locate the “Get Permission” button below the title description. This button is linked directly to the
title’s permission page on copyright.com. Alternatively, you can visit copyright.com and search by
title, ISBN, or ISSN.
For further questions about using the service on copyright.com, please contact:
Copyright Clearance Center
Phone: +1-(978) 750-8400 Fax: +1-(978) 750-4470 E-mail: info@copyright.com.
Independent verification should be sought for any data, advice or recommendations contained in this
book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to
persons or property arising from any methods, products, instructions, ideas or otherwise contained in
this publication.
This publication is designed to provide accurate and authoritative information with regard to the subject
matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in
rendering legal or any other professional services. If legal or any other expert assistance is required, the
services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS
JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A
COMMITTEE OF PUBLISHERS.
Additional color graphics may be available in the e-book version of this book.
Preface vii
Chapter 1 The Application of Parallel Monte Carlo Simulation
in the Reliability Evaluation of Composite Electric
Power Systems 1
Tran Ky Phuc, Pham Hai Binh and Le Thi Thanh Ha
Chapter 2 PK/PD Analysis of Stavudine-Gold Nanoparticles
by Monte Carlo Simulation 35
H. Zazo, A. Martín-Suárez and J. M. Lanao
Chapter 3 Monte Carlo Simulation of Magnetic Properties of
a Mixed Spins Ferrimagnetic and Kagomé Lattice
with RKKY Interaction 55
R. Masrour and A. Jabar
Bibliography 79
Related Nova Publication 149
Index 151
PREFACE
Chapter 1
ABSTRACT
Power system reliability is of critical importance to the electricity
sector. This paper presents a study on application of Monte Carlo
simulation in reliability assessment of composite electric power systems.
A distributed computing environment is used to reduce the computational
efforts. Several computer system architectures are considered to comply
with both technical and economical aspects. Programming framework
based on master-slave model is selected. The approach to partition of
simulation job and selection of distributed processing grain (yearly,
hourly) is considered in the paper. A suitable methodology for random
number generation in a distributed computing environment is presented.
For the purpose of the work, a computing tool based on a cluster of
multithread and multicore personal computers as hardware and Message
Passing Application Programming Interface as software framework is
Corresponding Author Email: trankyp@hotmail.com.
2 Tran Ky Phuc, Pham Hai Binh and Le Thi Thanh Ha
1. INTRODUCTION
At present, for most power systems, growing energy demand and system
operation in power market have caused operating conditions close to the
physical limits of system components. Therefore an effective computational
tool that allows evaluation of the power system reliability becomes a necessity.
For power system reliability evaluation, as well as analytical approaches,
the Monte Carlo Simulation (MCS) is one of the most important methods due
to its advantages (Billinton & Allan, 1994). Literature survey has shown that
the reliability evaluation based on MCS has undergone a great deal of
development in the last some decades (Borges, et al., 2001; Lingfeng Wang &
Singh, 2009) (Nisha R. Godha, et al., 2012). However, large actual power
systems and strict requirements on more accurate mathematical models of
components and systems (for example, using AC load flow and AC-optimal
power flow (OPF) techniques, statistical model for hydro power generation or
multistate outage model for transmission lines and generation) have often
caused unacceptable computation problem.
Variance reduction technique (Oliveira G.C., 1989), genetic algorithm
(GA) (Lingfeng Wang & Singh, 2009), artificial neural networks (ANN)
(Singh, et al., 2006) and parallel computing (Gubbala & Singh, 1995) are some
well-known methods applied for substantial improvement of computing
efficiency. Unfortunately, the intelligent methods like GA or ANN are
basically useful for non-sequential MCS only. For sequential MCS parallel
computing (Gubbala & Singh, 1995) (Borges, et al., 2001) (Tran Ky Phuc, et
al., 2011) is likely most suitable.
The Application of Parallel Monte Carlo Simulation … 3
There are two approaches for Monte Carlo simulation: state sampling (or
“non-sequential”) and sequential simulation (SMCS). In state sampling, the
system states are sampled randomly based on the probability distribution of
the components operating states. In sequential simulation, the chronological
behavior of the system is simulated by sampling sequences of system
operating states based on the probability distribution of the components states
duration. These sequences are sampled for several pre-defined periods, usually
years, in which case they are called yearly synthetic sequences.
The expected values of the main reliability indices, including frequency
and duration (F&D) indices can be calculated by both approaches. However,
estimates of specific energy supply interruption duration and the probability
distribution of duration related indices can only be determined by the
sequential simulation approach. Unfortunately, sequential simulation demands
a much higher computational effort, what may turn infeasible the analysis of
large systems on conventional computer platforms.
Some techniques have previously been developed to reduce the
computational resources necessary for MCS in order to achieve a more
efficient and timely convergence while continuing to accurately assess the
reliability of a given system. These methods can be classified into three
groups. The first group is based on variance reduction using statistical tools
(Oliveira G.C., 1989). The second group includes intelligent methods like
Genetic Algorithm (GA) (Lingfeng Wang & Singh, 2009), Artificial Neural
Networks (ANN) (Singh C., 1997) (Singh, et al., 2006) that can reduce state
space. The third group uses high performance computing (HPC) like parallel
processing to reduce computational time.
Notes should be made, that state space pruning with intelligent methods
like GA, ANN... is applied only for non-sequential MCS. Therefore, for
sequential MCS the parallel computing is more suitable approach for system
reliability evaluation.
NS
ni ( S )
( S j ,i )
E ( )
i 1 j 1
(1)
NS
V ( ) / NS (2)
E ( )
where
Ni,k : Number of interruptions occurring in year i, at Bus k
dj,i,k: Duration of the j-th interruption (hours) in year i at Bus k,
NS: Number of simulation years
BusENSj,i,k : Energy not supplied in MWh for the jth interruption, in year i
at Bus k
Ni : Number of system interruptions in year i,
dj,i : Duration of the jth system interruption (hours), in year i.
SysENSj,i: System energy not supplied in MWh for the jth interruption,
in year i.
8 Tran Ky Phuc, Pham Hai Binh and Le Thi Thanh Ha
Start
MASTER NODE
EDLC(k)=0; EENS(k)=0;EFLC(k)=0;
K=1,NBAll
EDLCS(k)=0; EENSS(k)=0;EFLCS(k)=0;
K=1,NB
NS=0
NS=NS+1
EENSSysY(NS)=0
ChronoUnitBranches(Cg,lg,mg,Cl,ll,ml,Cycle,CGgi,CLli) K=1,NB
Print EDLCK(k),EENSK(k),EFLCK(k)
LoadModel=Chrono
Yes PLCK(k)=EDLCK(k)/8760
ChronoHoursLoad(Pann, Pw%,Pd%,Ph%,PHi)
EDLCS(k)=EDLCS(k)+EDLCK(k)
No
NonChronoLoad(Pann) EENSS(k)=EENSS(k)+EENSK(k)
PLCS(k)=EDLCS(k)/8760
EDLCK(k)=0; k=1,NB
EFLCS(k)=EFLCS(k)+EFLCK(k)
EENSK(k)=0; k=1,NB
EENSSysY(NS)=EENSSysY(NS)+EENSK(k)
EFLCK(k)=0; k=1,NB
i=1,8760
COMPUTING NODE
K=1,NB
MASTER NODE
No EDLC(k)=EDLCS(k)/NS
OPF(i)
EENS(k)=EENSS(k)/NS
LoadCurtailment?
EFLC(k)=EFLCS(k)/NS
Yes
PLC(k)=EDLC(k)/8760
K=1,NB
No
LoadCurtailment at k-bus?
EDLCSys=0; EENSSys=0;EFLCSys=0
Yes K=1,NBAll
Yes
EDLCK(k)=EDLCK(k)+1
EDLCSys=EDLCSys+EDLC(k)
EENSSys=EENSSys+EENS(k)
EFLCSys=EFLCSys+EFLC(k)
EENSK(k)=EENSK(k)+DC(k)*Dt
PLCSys=EDLCSys/8760
CV(EENSSys,NS)
No LoadCurtailment at k-bus
at (i-1)-th time step?
(CV(EENSSys)<5%) OR
No (NS=NSmax)?
EFLCK(k)=EFLCK(k)+1 Yes
Stop
Notes should be made, that from the algorithm the loops for hourly and
yearly simulation (with PF and OPF solutions) consume large computation
time that need to be reduced by parallel processing.
For composite power system reliability evaluation in parallel computing
environment it is necessary to consider following issues: parallel computer
architecture, parallel programming languages and tools, parallel random
generator and parallel algorithm (including task partitioning).
The length of the repeated cycle is called the period of the generator. A
good generator is one with a long period and no discernible correlation
between elements of the sequence.
PRNGs for generating uniform distribution random numbers can be
classified into two groups, according to the basic arithmetic operations
utilized: RNGs based on modulo arithmetic and RNGs based on binary
arithmetic.
Modular Arithmetic Based RNGs.
RNGs of this type yield sequences of random numbers s by means of
linear recurrence modulo m, where m is a large integer.
Linear Congruential Generator (LCG): LCG is one of the well-known
random number generators that is defined by the following recurrence relation:
where s is the sequence of the generated random numbers and m > 0, 0 <a <
m, and 0 c; s0 < m.
If uniform distribution on [0; 1) is needed, then u = s/m should be used as
the output sequence. Numbers a, c and m should be carefully chosen to make
sure that maximum period can be achieved. LCG can be easily implemented
on computer hardware which can provide modulo arithmetic by storage-bit
truncation. RNG using LCG is provided with C library (rand()) as well as
many other languages. LCG has a relatively short period (at most 232 for 32-bit
integer) compared to other more complicated ones.
Multiple Recursive Generator (MRG): MRG is a derivative of LCG and
can achieve much longer period. A MRG of order k is defined as follows:
The recurrence has maximal period length mk-1, if tuple (a1,..., ak) has
certain properties (Knuth, 1969).
Combined Multiple Recursive Generator (CMR): CMR combines multiple
MRGs and can obtain better statistical properties and longer periods compared
with a single MRG. A well-known implementation of CMR, CMR32k3a
(L’Ecuyer, 1999), combines two MRGs:
The Application of Parallel Monte Carlo Simulation … 17
𝑠𝑖 = 𝑥𝑖 − 𝑦𝑖 mod 𝑚1 (5)
operations.
The underlying algorithm of a xorshift with four seeds (x; y; z; s) is as
follows:
𝑡 = (𝑥(𝑥𝑖 ≪ 𝑎))
𝑥=𝑦
𝑦=𝑧
𝑧=𝑠
0 𝐼𝑤−1
𝐴=( ),
𝑎𝑤 𝑎𝑤−1 𝑎1
where Iw-1 is the (w -1) x (w - 1) identity matrix and ai, i = 1,..., w take values
of either 0 or 1.
Define
(𝑤:(𝑙+1)) (𝑙:1)
𝑥𝑖+𝑟 = (𝑥𝑖+𝑟 (𝑥𝑖 | 𝑥𝑖+1 ) 𝐴),
𝑧 = 𝑥𝑖+𝑟 (𝑥𝑖+𝑟 ≥ 𝑡1 )
𝑧 = 𝑧 (𝑥 ≫ 𝑡4 )
where t1, t2, t3 and t4 are integers and m1 and m2 are bit-masks and ‘&’ is a bit-
wise AND operation.
si+r; i = 1, 2,... form the required sequence on interval (0; 1].
With proper parameter values, MT can generate sequence with a period as
long as 219,937 and extremely good statistical properties(Matsumoto & Takuji
Nishimura, 1998). Strategies for selecting good initial values are studied in
(Mutsuo Saito & Makoto Matsumoto, 2008) while Saito et al. proposed
efficient implementation for fast execution on GPUs.
(David Hill, et al., 2012), according to that the techniques can be classified
to some main groups: Partitioning a main sequence (a unique original stream)
and Partitioning multiple streams via parametrization.
Partitioning a unique original stream is the approach which splits random
numbers from the considered cycle between the different processing elements
(PEs). Depending on the method that splits the main stream, this approach can
be classified into four main techniques.
1. Central server technique: It consists in using a central sever, running an
RNG and providing on-demand pseudo-random numbers to different PEs. The
advantage of this technique is avoiding the problem of generating multiple
independent random sequences.” However, this approach makes
reproducibility hard to achieve because of providing on-demand that might
change the order in which numbers will be distributed to PEs. Another
drawback is that a bottleneck will be created if too many PEs are applied. As a
result, this technique is not quite suitable for scientific applications.
2. Sequence splitting: It consists in allocating non-overlapping, contiguous
and equally sized blocks from the original random stream to form substream.
When partitioning a sequence xi, i =0, 1,... into N substreams, the j-th
substream is xk+(j−1)m, k =0,..., m−1, where m is the length of each substream.
One of the drawbacks is that m must be chosen so that each substream is long
enough to achieve the stochastic simulation performed by the corresponding
process. In addition, long-range correlations in the initial RNG can lead to
small-range correlations between the potential substreams.
3. Random spacing (indexed sequences methods): The random spacing or
indexed sequences method builds a partition of n streams by initializing the
same generator with n random statuses. In the case of old LCGs, it was named
random seeding. For modern generators with a more complex status, the
random statuses are generated with another RNG, and this technique is
interesting when generators have a huge period. This technique should be
avoided when the period of the RNG is smaller than 260 and when a single
simulation is consuming more than 1011 random numbers. Random spacing is
suitable for limited-memory environments like GPUs.
4. Leap frog: The leap frog (LF) is the way to partition a random stream in
the manner of dealing cards to several players. Random numbers are allocated
in turn to PEs, in the same way as cards are dealt to players. Pragmatically, let
each processor hold an i identifier. Every such PE will build a Yi substream
from an X original random stream such as Yi = Xi, Xi+N,..., Xi+kN, with N equal
to the number of processors(Aluru S., 1997).
20 Tran Ky Phuc, Pham Hai Binh and Le Thi Thanh Ha
Given the period P of the global sequence, the period of each stream is
P/N. As with the splitting technique, the long-range correlations in the initial
RNG can lead to small-range correlations between the potential substreams,
particularly if we have a large number of PEs. This technique needs to be used
with caution depending on the environment it is set up in. Thus, to preserve
reproducibility between two executions of the same simulation, one should not
use this technique when the parallelism grain is not fixed yet.
Partitioning multiple streams with parameterization: This approach
consists in using several declinations of the same PRNG: each generator has
the same structure and generation mechanism with a unique parameter set.
Although no mathematical proof can establish this independence, some
implementations of parameterization are safe according to the current state of
the art.
The use of random numbers in parallel Monte Carlo simulation is still a
challenging technical problem. Parallel simulations must first consider the
choice of a nice sequential generator. Test batteries are the only empirical way
to ensure that the generator in use fits for s wide set of applications. Selection
of techniques for parallel RNDs (the sequence splitting, the random spacing,
the leap frog or the parametrization...) largely depends on the nature of
application, the partitioning method and the results of current test methods
when available.
4. CASE STUDY
This section describes some issues related to implementing a VietReli
called computing tool for composite power system reliability evaluation based
on Monte Carlo simulation and parallel processing.
Master Node
User Interface
MPAPI
Start simulation
alt
The tests are explored in several power systems, mainly the 24 bus IEEE
Reliability Test System (IEEE Committee Report, 1979) and VietPS – the
Vietnam power system. The communication environment is a Local Area
Network with up to 8 PCs, each PC with core i5 processor. The LAN speed is
100 Mbps. The parallel and distributed processing is implemented in Visual
C# with TCP/IP protocol for message passing based on a MPAPI framework
(Thomsen, 2008).
The testing procedure is described as follows. First, the test systems are
examined for reliability evaluation on standalone multicore PC (core i5) using
different numbers of threads (from 1 to 8). Secondly, the test systems are
The Application of Parallel Monte Carlo Simulation … 25
CONCLUSION
MCS based composite system reliability evaluation has often faced time-
consuming computation problem, in particular for large actual power systems
and accurate mathematical models of system components. This problem
practically cannot be solved without using intelligent techniques especially
parallel computing environment. As IT and parallel computing grows and
changes rapidly, the methodology for reliability evaluation should be
systematically and frequently approached with updating new IT tools and
methods.
Many surveys have shown that at present all popular parallel computers
are based on MIMD architecture. MIMD computers can be classified into
shared memory machines - SMMs (or parallel computing systems) and
distributed memory machines - DMMs (or distributed computing systems).
Nowadays High Performance Cluster (i.e., Supercomputers); Cluster of
Multicore PCs and Network of Workstations are the parallel computer
architectures options to be considered. Final selection of the computer
architecture depends on the required performance and available budget.
OpenMP and MPI are the most popular programming tools and suitable
for SMM and DMM respectively. For hybrid shared memory and distributed
memory architecture both OpenMP and MPI languages should be used. There
are several parallel programming tools that combined OpenMP and MPI
functions together.
Random number generation is the first critical issue for parallel MCS
based composite system reliability evaluation. Partitioning a unique original
stream (which includes central server technique, sequence splitting, random
spacing and leap frog) and partitioning multiple streams with parameterization
are two groups of well-respected methods that should be considered. Basic
RNGs can be modular arithmetic based or binary arithmetic based algorithms
depending on size of the simulated system and result of statistical tests for
RNGs.
Parallel algorithm including task partitioning is the second critical issue.
The performance of a parallel algorithm largely depends on whether the
problem is fine-grained, coarse-grained or embarrassingly parallel. For the
MCS based reliability evaluation a grain can be an hour, several hours or a
year of simulation. Selection of suitable grain depends on largely computer
architecture and nature of the problem.
In the case study a cluster of multicore PCs was selected for two reasons.
First, the cluster is available and cost-effective. Second, cluster of multicore
The Application of Parallel Monte Carlo Simulation … 27
ACKNOWLEDGMENTS
The work presented in this chapter was supported by the Ministry of
Science and Technology and Ministry of Industry and Trade Grant
44/2012G/HD-ĐTĐL.
REFERENCES
Aluru S., 1997. Lagged fibonacci random number generators for distributed
memory parallel computers. Journal of Parallel and Distributed
Computing, Volume 45(1), pp. 1-12.
Billinton, R. & Allan, R., 1994. Reliability evaluation of power systems.
Second ed. New York: Plenum Press.
Billinton, R. & Wangdee W, 2006. Predicting bulk electricity system
reliability performance indices using sequential Monte Carlo simulation.
IEEE Transactions on Power Delivery, March, 21(2), pp. 909 - 917.
Boggan, Sha'Kia & Daniel M. Pressel, 2007. GPUs: An Emerging Platform
for General-Purpose Computation, s.l.: U. S. Army Research Lab..
28 Tran Ky Phuc, Pham Hai Binh and Le Thi Thanh Ha
Borges, C. L. T., Falcao, D. M., Mello, J. & Melo, A., 2001. Composite
Reliability Evaluation by Sequential Monte Carlo Simulation on Parallel
and Distributed Processing Environments. IEEE Trans. on Power Systems,
May, 16(2), pp. 203-209.
Buyaa, R., 1999. High Performance Cluster Computing:,,,. s.l.:s.n.
Chee, S. Y. & Buyya, R., 2004. Cluster Computing: High-Performance, High-
Avalability, and High-Throughput Processing on a Network of Computers,
s.l.: s.n.
Chee, S. Y. & Buyya, R., 2004. Cluster Computing: High-Performance, High-
Avalability, and High-Throughput Processing on a Network of Computers,
s.l.: s.n.
Creel M., G. W., 2007. Multi-core CPUs, Clusters, and Grid Computing: A
Tutorial, NY: s.n.
David Hill, Claude Mazel, Jonathan Passerat-Palmba & Mamadou Kaba
Traore, 2012. Distributionof Random Streams for Simulation
Practitioners.. In: Concurrency and Computation: Practice and
Experience. s.l.:Wiley, pp. 1427 - 1442.
Deng Z., S. C., 1992. A new approach to reliability evaluation of
interconnected power systems including planned outages and f. IEEE
Trans. PWRS, 7(2), pp. 734-743.
Endrenyi, J., 1979. Reliability modeling in electric power systems. New York:
Wiley & Sons.
EPRI, 1987. Composite System Reliability Evaluation Methods, Final Report
on Research Project 2473-10, EPRI EL-5178, s.l.: s.n.
Flynn, M. J., 1972. Some Computer Organizations and Their Effectiveness.
IEEE Trans. Comput. C–21 (9):, Volume C–21 (9), p. 948–960.
G. Jost, H. Jin, D. Mey & F. Hatay, 2003. Comparing the OpenMP, MPI, and
Hybrid Programming Paradigms on an SMP Cluster. Fifth European
Workshop on OpenMP (EWOMP03). s.l., s.n.
Gottlieb, Allan & Almasi, George S, 1989. Highly parallel computing..
Redwood City, Calif.: Benjamin/Cummings.
Gubbala, N. & Singh, C., 1995. Models and considerations for parallel
implementation of Monte Carlo simulation methods for power system
reliability evaluation. May, 10(2), pp. 779-.
Guo Liang Chen & Yun Quan Zhang, Sept. 2012. Survey on Parallel
Computing. Journal of Computer Science & Technology, 21(5), pp. 665-
673.
Hawick, 2010. Mixing muli-Core CPUs and GPUs. Res. Lett. Inf. Math. Sci.,
Volume 14.
The Application of Parallel Monte Carlo Simulation … 29
Singh C., M. J., 1997. Composite system Reliability Evaluation using State
Space Pruning. IEEE Transactions on Power Systems, 12(1).
Singh, C., Xiaochuan Luo & Hyu, 2006. Power System Adequacy and Security
Calculations Using Monte Carlo Simulation incorporating Intelligent
System Methodology. Stockholm, Sweden, KTH.
Thomsen, F., 2008. Introduction to Message Passing and the MPAPI
framework. [Online].
Tikhonov, A. N., 1963. Solution of incorrectly formulated problems and the
regularization. Soviet Math. Dokl. 4, p. 1035–1038.
Tran Ky Phuc, Vu Toan Thang & Le Thi Thanh Ha, 2011. A Monte Carlo
simulation based computing tool for assessing reliability of generation
systems. Journal of science & technology - Technical universities, Issue
85, pp. 7-12.
Tran Ky Phuc, Vu Toan Thang, Pham Hai Binh & Le Thi Thanh Ha, 2015. A
Computing Tools for Composite Power System Reliability Evaluation
Based on Monte Carlo Simulation and Parallel Processing: Proceedings
of the Second Int. Conf. on Mechatronics and Automatic Control. s.l.,
Verlag Springer.
Vietnam Institute of Energy, 2011, 2014. The National Master Plan for power
development for the 2011 – 2020 period with the vision to 2030, Hanoi:
Vietnam Institute of Energy.
Website, 2012. http://www.kurzweilai.net/ibm-announces-20-petaflops-
supercomputer. [Online].
Website, n.d. http://www.10gea.org, s.l.: s.n.
Website, n.d. http://www.mcs.anl.gov/research/ projects/mpi. [Online].
Website, n.d. https://en.wikipedia.org/wiki/Parallel_computing#cite_note-1.
[Online].
Website, n.d. Wikipedia. [Online].
BIOGRAPHICAL SKETCHES
Pham Hai Binh
Professional Appointments:
2010-2012: Researcher at International Research Institute MICA
2012-2016: Senior Software Engineer / ECM Technical Leader at
VietSoftware International Inc.
Le Thi Thanh Ha
Professional Appointments:
Since 1998: Teacher in Hanoi University of Technology
Tran Ky Phuc
Professional Appointments:
Before 2000: Researcher in Electricity of Vietnam.
34 Tran Ky Phuc, Pham Hai Binh and Le Thi Thanh Ha
Chapter 2
PK/PD ANALYSIS OF
STAVUDINE-GOLD NANOPARTICLES
BY MONTE CARLO SIMULATION
ABSTRACT
Introduction: Antiretroviral drug treatments allow for the control of
HIV infection keeping low levels of viral load during various years. After
that time, viral load increases again due to the virus replication from
reservoirs, such as macrophages, where drugs do not reach successful
therapeutic conditions. The use of delivery systems such as gold
nanoparticles (AuNPs), helps to target the drug inside both HIV target
and reservoir cells. Different pharmacokinetic/pharmacodynamic
(PK/PD) models have been developed to study drug plasma and tissue
levels, as well as the pharmacological response in HIV infection. The aim
of this work is the development of a PK/PD model to evaluate, by Monte
Carlo simulation as a data maximization strategy, the antiviral activity of
two stavudine formulations: conventional stavudine and stavudine-gold
nanoparticles (stavudine-AuNPs).
36 H. Zazo, A. Martín-Suárez and J. M. Lanao
1. INTRODUCTION
Antiretroviral drug treatments allow for the control of human
immunodeficiency virus (HIV) infection keeping low levels of viral load
during various years. After that time, viral load increases again due to the virus
replication from reservoirs, like macrophages (Cassol, Alfano, Biswas, & Poli,
2006a), where the drug does not reach successfully. Mononuclear phagocytes
have been proposed as the main reservoir for viral dissemination, reaching
tissues with limited drug entry (Shegokar & Singh, 2011).
PK/PD Analysis of Stavudine-Gold Nanoparticles … 37
2. METHODS
2.1. Dosage Regimens Design
escape routes. The IC50 of stavudine has a huge variability both in T cells and
macrophage. The CV for T cells was 90 % and 75 % for macrophages based
on the values of the literature (Bristol-Myers Squibb Company, n.d.; Dolin et
al., 2007; Shegokar & Singh, 2011).
Free stavudine inside macrophages inhibited the viral load from them
(INHM), while the free stavudine serum inhibited viral load from T cells
(INH). The differential equations used are:
A1 ivFreeDrugSerum A2 ivFreeDrugIntraCell
400
1.4
1.2 300
1.0
0.8 200
0.6
0.4 100
0.2
0.0 0
100.0 100.2 100.4 100.6 100.8 101.0 101.2 101.4 101.6 101.8 100.0 100.2 100.4 100.6 100.8 101.0 101.2 101.4 101.6 101.8
Time (day) Time (day)
B1 OralFreeDrugSerum B2 OralFreeDrugIntraCell
400
1.4
intracellular concentration (mg/L)
Serum concentration (mg/L)
1.2
300
1.0
0.8 200
0.6
0.4 100
0.2
0.0 0
100.0 100.2 100.4 100.6 100.8 101.0 101.2 101.4 101.6 101.8 100.0 100.2 100.4 100.6 100.8 101.0 101.2 101.4 101.6 101.8
Time (day) Time (day)
C1 GNPFreeDrugSerum
C2 GNPFreeDrugIntraCell
400
1.4
intracellular concentration (mg/L)
serum concentration (mg/L)
1.2 300
1.0
0.8 200
0.6
0.4 100
0.2
0.0 0
100.0 100.2 100.4 100.6 100.8 101.0 101.2 101.4 101.6 101.8 100.0 100.2 100.4 100.6 100.8 101.0 101.2 101.4 101.6 101.8
Time (day) Time (day)
Figure 3. Monte Carlo Simulation of stavudine concentrations for (A) conventional i.v.
stavudine (B) conventional oral stavudine and (C) stavudine-AuNPs i.v. administration
at the dose of 40 mg/12h (Serum concentrations [1] and intracellular macrophages
concentrations [2]).
PK/PD Analysis of Stavudine-Gold Nanoparticles … 45
Figure 4. Probability of exceeding a specific viral load level after 10 years with
different treatments.
0.008
0.007
Probability Density
0.006
0.005
0.004
0.003
0.002
0.001
0.000
100 200 300 400
Macrophages [cell/uL]
CONCLUSION
Modelling and simulation were used to interpret preclinical data. We have
been able to characterize the dose-response curves of viral load and
immunological cell levels with a new HIV therapy, much faster and cheaper
than is expected with clinical trials.
According to our PK/PD model, the use of nanoparticles treatment results
in a better control of the infection. This treatment:
Once preclinical and clinical data became available, the model will be
updated and used to make predictions with higher confidence in the following
phases of the studies. This model-based approach could be used as a tool to
communicate with the clinical team, making the risks and assumptions clearer,
given the knowledge from the model at that time.
GLOSSARY
HIV Human immunodeficiency virus
AuNPs Gold nanoparticles
PK/PD Pharmacokinetic and Pharmacodynamic
PTA Probability of Target Attainment
SC Stavudine serum concentration
MC Stavudine intracellular macrophage concentration
NP-SC Stavudine-AuNP serum concentration
NP-MC Stavudine-AuNP intracellular macrophage concentration
Drug-SC Free stavudine from stavudine-AuNP serum concentration
50 H. Zazo, A. Martín-Suárez and J. M. Lanao
REFERENCES
Averill M.L., K. W. . (2000). Simulation Modeling and Analysis, (Third
Edition). McGraw-Hill. Retrieved from http://www.mhhe.com/
engcs/industrial/lawkelton/.
Baggaley, R. F., Powers, K. A., & Boily, M.-C. (2011). What do mathematical
models tell us about the emergence and spread of drug-resistant HIV?
Current Opinion in HIV and AIDS, 6(2), 131–140. http://doi.org/10.1097/
COH.0b013e328343ad03.
Bonhoeffer, S. (1998). Models of viral kinetics and drug resistance in HIV-1
infection. AIDS Patient Care and STDs, 12(10), 769–774. http://doi.org/
10.1089/apc.1998.12.769.
Bonhoeffer, S., & Nowak, M. A. (1997). Pre-existence and emergence of drug
resistance in HIV-1 infection. Proceedings. Biological Sciences/The Royal
Society, 264(1382), 631–637. http://doi.org/10.1098/rspb.1997.0089.
Bristol-Myers Squibb Company. (n.d.). ZERIT®XR (stavudine). U.S. Food an
d Drug Administration. Retrieved from http://www.
accessdata.fda.gov/drugsatfda_docs/label/2005/21453s004lbl.pdf.
Burg, D., Rong, L., Neumann, A. U., & Dahari, H. (2009). Mathematical
modeling of viral kinetics under immune control during primary HIV-1
infection. Journal of Theoretical Biology, 259(4), 751–759.
http://doi.org/10.1016/j.jtbi.2009.04.010.
Cassol, E., Alfano, M., Biswas, P., & Poli, G. (2006a). Monocyte-derived
macrophages and myeloid cell lines as targets of HIV-1 replication and
persistence. Journal of Leukocyte Biology, 80(5), 1018–1030.
http://doi.org/10.1189/jlb.0306150.
Cassol, E., Alfano, M., Biswas, P., & Poli, G. (2006b). Monocyte-derived
macrophages and myeloid cell lines as targets of HIV-1 replication and
persistence. Journal of Leukocyte Biology, 80(5), 1018–1030.
http://doi.org/10.1189/jlb.0306150.
Chellat, F., Merhi, Y., Moreau, A., & Yahia, L. (2005). Therapeutic potential
of nanoparticulate systems for macrophage targeting. Biomaterials,
26(35), 7260–7275. http://doi.org/10.1016/j.biomaterials.2005.05.044.
Dolin, R., Masur, H., & Saag, M. S. (2007). AIDS Therapy. Elsevier Health
Sciences.
Duffin, R. P., & Tullis, R. H. (2002). Mathematical models of the complete
course of HIV infection and AIDS. Journal of Theoretical Medicine, 4(4),
215–221.
52 H. Zazo, A. Martín-Suárez and J. M. Lanao
Garg, M., Asthana, A., Agashe, H. B., Agrawal, G. P., & Jain, N. K. (2006).
Stavudine-loaded mannosylated liposomes: in-vitro anti-HIV-I activity,
tissue distribution and pharmacokinetics. The Journal of Pharmacy and
Pharmacology, 58(5), 605–616. http://doi.org/10.1211/jpp.58.5.0005.
Iman, R.L., J.C. Helton, J.E., Campbell, & J.E. (1981). An Approach to
Sensitivity Analysis of Computer Models, Part 1. Introduction, Input
Variable Sele ction and Preliminary Variable Assessment. Journal of
Quality Technology, 13(3), 174–183.
Kepler, T. B., & Perelson, A. S. (1998). Drug concentration heterogeneity
facilitates the evolution of drug resistance. Proceedings of the National
Academy of Sciences of the United States of America, 95(20), 11514–
11519.
Koizumi, Y., & Iwami, S. (2014). Mathematical modeling of multi-drugs
therapy: a challenge for determining the optimal combinations of antiviral
drugs. Theoretical Biology & Medical Modelling, 11, 41.
http://doi.org/10.1186/1742-4682-11-41.
Levy, J. A. (2007). HIV and the Pathogenesis of AIDS. ASM Press.
Li, M., Al-Jamal, K. T., Kostarelos, K., & Reineke, J. (2010). Physiologically
based pharmacokinetic modeling of nanoparticles. ACS Nano, 4(11),
6303–6317. http://doi.org/10.1021/nn1018818.
Maderuelo, C., Zarzuelo, A., & Lanao, J. M. (2011). Optimization of release
kinetics from sustained-release formulations using model-independent
pharmacokinetic simulation. Journal of Pharmaceutical Sciences, 100(8),
3260–3267. http://doi.org/10.1002/jps.22565
Martín, J. C., & Bandrés, J. C. (1999). Cells of the monocyte-macrophage
lineage and pathogenesis of HIV-1 infection. Journal of Acquired Immune
Deficiency Syndromes (1999), 22(5), 413–429.
Nestorov, I. (2001). Modelling and simulation of variability and uncertainty in
toxicokinetics and pharmacokinetics. Toxicology Letters, 120(1-3), 411–
420.
Nowak, M. A., & Bangham, C. R. M. (1996). Population Dynamics of
Immune Responses to Persistent Viruses. Science, 272(5258), 74–79.
http://doi.org/10.1126/science.272.5258.74.
Perelson, A. S. (2002). Modelling viral and immune system dynamics. Nature
Reviews. Immunology, 2(1), 28–36. http://doi.org/10.1038/nri700.
Raimundo, S. M., Yang, H. M., Venturino, E., & Massad, E. (2012). Modeling
the emergence of HIV-1 drug resistance resulting from antiretroviral
therapy: insights from theoretical and numerical studies. Bio Systems,
108(1-3), 1–13. http://doi.org/10.1016/j.biosystems.2011.11.009.
PK/PD Analysis of Stavudine-Gold Nanoparticles … 53
Chapter 3
ABSTRACT
The magnetic properties of the kagomé lattice have been studied with
Ruderman–Kittel–Kasuya–Yosida (RKKY) exchange interactions in a
spin-7/2 and alternate mixed spin-5/2 and spin-2 Ising model on the
Bethe lattice have been studied by using the Monte Carlo simulations.
The RKKY interaction between the two magnetic layers is considered for
different distances. The magnetizations and magnetic susceptibilities of
this lattice are given for different triquadratic interactions around each
triangular face. The ground state phase diagrams of kagomé lattice and
alternate mixed spin-5/2 and spin-2 Ising model on the Bethe lattice have
been obtained. The thermal total magnetization and magnetization of
spins-5/2 and spin-2 with the different exchange interactions, external
magnetic field and temperatures have been studied. The critical
Corresponding author Email: rachidmasrour@hotmail.com.
56 R. Masrour and A. Jabar
temperature has been obtained for two systems. The magnetic hysteresis
cycle of kagomé lattice with RKKY interactions and on the Bethe lattice
are obtained for different temperatures and for different crystal field with
a fixed size.
1. INTRODUCTION
Many works have been carried out to investigate the magnetic properties
of the mixed spin-2 and spin-5/2 ferrimagntic system with different numerical
methods, such as Monte Carlo simulations (MCS) [1–5], effective-field theory
(EFT) [6–10], the exact recursion equations [11], mean field theory (MFT)
[12–14] and the Green’s function approach [15–17]. These studies have
focused mainly on magnetic and thermodynamic properties of the mixed spin
ferrimagnetic system, including magnetization, the initial susceptibility,
internal energy, specific heat and the ground phase diagram. Especially, the
tricritical and compensation behaviors have been also investigated in detail [6,
8, 11, 12]. The ground state phase diagrams and the critical temperature and
compensation temperature have been deduced of alternate mixed spin-5/2 and
spin-2 Ising model on the Bethe lattice has been obtained [18] and [19, 20],
respectively. The thermal magnetizations, the magnetic susceptibilities and the
transition temperature of the bilayer spin-1 and 2 on the Bethe lattice are
studied for different values of crystal field and intralayer coupling constants of
the two layers and interlayer coupling constant between the layers [21]. The
magnetic properties of the kagomé lattice have been studied with Ruderman–
Kittel–Kasuya–Yosida exchange interactions in a spin-7/2 Ising model using
MCS [22]. Monte Carlo simulation has been used to study the magnetic
properties of a mixed spin-1 and spin-3/2 ferrimagnetic through Ising model
on a honeycomb lattice [23] and on a ferrimagnetic nanoparticles [24]. The
thermal and magnetic hysteresis cycles of a mixed spins (2-1) hexagonal Ising
nanowire with core–shell structure are given for different values of the crystal
field using MCS by Ref. [25]. The dynamic hysteresis behaviors for the two-
dimensional (2D) mixed spin (2, 5/2) ferrimagnetic Ising model on a
hexagonal lattice in an oscillating magnetic field within the framework of
dynamic mean-field calculations is investigated [26]. The effect of inter-
ladder ferromagnetic (FM) coupling in spin-1/2 two-leg ladders with
Monte Carlo Simulation of Magnetic Properties … 57
where i, j stand for the first nearest neighbor sites (i and j) and i, j , k is
over all triplets of sites belonging to elementary triangles. D represent the
crystal field and h is the external magnetic field. JRKKY is the Ruderman-
Kittel-Kasuya-Yoshida (RKKY) interaction between the two magnetic layers.
The spin moment is: S = ±7/2;±5/2;±3/2; ±1/2. The new parameters used in the
text are: R3 = J3/J and J = 1.0. To better understand the physical origin of the
exchange coupling phenomenon indirect, several theoretical models have been
developed, with the first model it was applied it to Ruderman Kittel Kasuya
and Yoschida, this model Consider a uniform spin distribution within each
layer ferromagnetic coupling between two remote JRKKY spin plans is d such as
given in Figure 1 of the form:
a2 J0
J RKKY cos k f d
d2 (2)
between the two magnetic layers in the system such as given in Figure 1.
Where Kf = 0.5 and a2J0 = 1. Kf is the Fermi level, J0 is a magnetic coupling
constant [31-33]. In the mixed spins-5/2 and 2 Ising model on the Bethe lattice
with different spins such as in see Figure 2, the Hamiltonian of given by Eq.(1)
includes nearest neighbors interactions, the crystal field and external magnetic
field is becomes:
H J
i , j
i j J S S
i , k
i k D i2 Sk2 h i Sk
i i
k k
(3)
where i, j stand for the first nearest neighbor spins i and j, D represent the
crystal field and h is the external magnetic field. The J and JS are the
exchange interactions between the first nearest-neighbor magnetic atoms with
spins - and S-, respectively. The spins moment S and are: ±5/2, ±3/2,
±1/2 and ±2, ±1, 0, respectively. In full text, the J has been taken 1.
Figure 2. Bethe lattice for mixed spin-5/2 and spin-2 Ising model on the Bethe with
N = 9 and NS = 15 spins numbers.
Our data were generated with 105 Monte Carlo steps per spin, discarding
the first 104 Monte Carlo simulations. Starting from different initial conditions,
we performed the average of each parameter and estimate the Monte Carlo
simulations, averaging over many initial conditions. Our program calculates
the following parameters, namely:
The magnetization of kagomé lattice is given by:
Monte Carlo Simulation of Magnetic Properties … 61
1
M
N
S
i
i
(4)
The magnetizations of each atoms with spins and S on Bethe lattice are:
1
M
N
i
i (5)
1
MS
NS
S i
i
(6)
15M S 9M
M
24 (7)
1 (8)
E H
N
where N = NS + N
The magnetic susceptibilities of atoms with spins and S are given by:
2 2
1 1 (9)
N
i
N
i
i i
2 2
1 1
Si Si (10)
N S i NS
S
i
The total susceptibility for mixed spins-S and Ising model on the Bethe
lattice mixed spins-S and Ising model on the Bethe lattice is:
9 15 S
tot
14 (11)
62 R. Masrour and A. Jabar
2
Si
2 (12)
Si
N 2 i i
1
where k T , T denotes the absolute temperature and k B is the Boltzmann’s
B
constant.
Figures 3a-c. The ground state phase diagrams of kagomé lattice with RKKY
interaction on the ((Δ/J1, R3) with h/J1 = 0.0), ((h/J1,R3), Δ/J1 = 0) with (h, Δ) and ((h/J1,
Δ/J1) with R3 = 1.0), respectively for L = 4.
Monte Carlo simulation. The ground state phase diagrams of kagomé lattice
with RKKY interaction on the ((Δ/J1, R3) (1) with h/J1 = 0.0), ((h/J1, R3) (2),
Δ/J1 = 0) with (h, Δ) and ((h/J1, Δ/J1) (3) with R3 = 1.0), respectively for L = 4
are presented in Figures 3a-3c. In plane (1), the phases obtained are: (+1),+7/2,
-1/2 and -7/2. For Δ/J1 = -4 and R3 = 0, the (+1),+7/2, -1/2 and -7/2 phases are
found. The (7/2) and (1/2) are symmetrical in relation to the R3 = 0 axis and
some either the value of the crystal field (see Figure 3a). When we varied R3
(with R3 < 0) versus the external field the (3/2) and (5/2) are appeared such
as given in Figure 3b. by against in the (3) plane, the appearance of phase
(7/2) and (5/2) are observed and the absence of the phase (3/2) and (1/2)
such as given in Figure 3c. Figures 4(a)-d, show the ground state phase
diagrams of alternate mixed spin-5/2 and spin-2 Ising model on the Bethe
lattice in (Δ, JS), (h, JS), (h, Δ) for JS = 1.0 and (h, Δ) planes for JS = -1.0,
respectively. The obtained configurations are presented in Figures 4(a)-(d) and
are given in table 1, where more than one phase can coexist. It should be
mentioned that the ground state phase diagram is important in classifying the
different phase regions of the model for the phase diagrams at higher
temperatures. This behavior is observed in previously work [35]. The I, II, V,
VI, XV, XVI, XVII, XVIII, XIX, XX, XXIII, XXIV phases represent the usual
ferromagnetic ordering. The VII, VIII, IX, X, XI, XII phases presents the
paramagnetic ordering. The others phases are presented in Table 1 with the
antiferromagnetic ordering. For h > 0 the ferromagnetic (FM) and
antiferromagnetic (AFM) phases are observed with JS > 0 and JS < 0,
respectively. The mixed spins (2, 5/2) Ising system displays the dynamic
reentrant behavior for the FM/FM, AFM/FM and AFM/AFM interactions. For
h < 0 the FM and AFM phases are observed with JS > 0 and JS < 0,
respectively (see Figure 4c). When I change the sign of the coupling between
spin- and spin-S, the nature and number of phase changes its change (see
Figures 4(d) and 4(e)). We have given in Figure 5(a)-b, the thermal
magnetization and magnetic susceptibility of mixed spins-5/2 and 2 Ising
model on the Bethe lattice for JS = 1.0 and JS = -1.0 with Δ = 0.0 and h =
0.75. The obtained transitions temperatures for JS > 0 and JS < 0 are 5 and
3.3, respectively. These values are comparable with those given by effective-
field theory and the Glauber-type stochastic dynamics approach [36, 37, 19].
The second-order phase transition temperature is observed in these figures.
This behavior is observed in the previously work [38]. In the previously study
[19] the temperature compensation and transition temperatures are observed
and in the present study we have obtained only the transition temperature. The
64 R. Masrour and A. Jabar
thermal total magnetization and magnetic susceptibility for JS = 1, 1.5, 2 and
JS = -1, -1.5, -2 are presented in Figures 6(a) and 6b, respectively with Δ = 0.0
and h = 0.75. The obtained values of transitions temperatures for JS = 1, 1.5, 2
are: 5.70, 7 and 8.1, respectively. The obtained values for JS = -1, -1.5, -2 are:
3.45, 4.3, 5.4, respectively. The transitions temperatures increases with
increasing the absolute value of exchange interactions JS. These values are
comparable with those given by Refs. [39, 40]. The obtained values for N = 24
spins in the non-alternate mixed spins [19] is inferior to those obtained in the
presented work. From the figure, one can see that the sublattice with spin-S is
more ordered than the sublattice with spin- below TC, i.e., |MS| > |M|. We
have presented in Figure 7, the thermal magnetization and magnetic
susceptibility of kagomé lattice for R3 = 0.1, 0.3, 0.5 with Δ/J = 0, h/J = 0 for L
= 4. We see that the magnetic moment remains in the high-spin state in the low
temperature range for all the structures and then drops to near zero at the Curie
temperature. These values indicate that the ferromagnetism that we observed
in these kagome lattices can be detected at room temperature. Thus, the
designer 2D ferromagnetic kagome lattices may have advantages over the well
known dilute magnetic semiconductors in the applications of spintronic
materials. The critical temperatures TC/J are obtained from the maximum of
the magnetic susceptibilities. The critical temperatures obtained values are:
16.2, 20, 24 for R3 = 0.1, 0.3 and 0.5, respectively. The critical temperatures
obtained increase with increasing the exchange interactions such as given in
Figure 8, for h/J = D/J = 0 and h/J = 0. This behavior is observed in previously
work [41]. The JRKKY exchange interactions as a function of distance between
the two magnetic layers have been shown in Figure 9. The intralayer RKKY
interactions of kagomé lattice have same oscillatory factor. The same behavior
is observed in previously work [42]. At very large impurity distances, JRKKY
shows a decay close to 1/d which is similar to the long-distance behavior of
the RKKY interaction in the ordinary two-dimensional electron gas [41].
Similar result for power law decay of JRKKY has been reported in Ref.[43].
Figures 10a-c show the variation of the magnetization versus the R3 for L = 1,
2, 3, 4, T/J = 5, Δ/J = 0.0, D/J = 0,-1, -1.2, -1.5, T/J = 5 and T/J = 4, 5, 8, 10,
Δ/J = 0, respectively with h/J = 0 and L = 4. The magnetization is independent
of the size effect for R3 < 0.2 and R3 > 0.3. The magnetization increase rapidly
until reached saturation for 0.2 < R3 < 0.3. The same behavior is observed in
the crystal field and temperature effect for R3 > 0.3. The magnetization versus
the crystal field D/J for (L = 1, 2, 3, 4, T/J = 1, R3 = 0.01), (R3 = 0.01, 0.1, 0.2,
0.4, T/J = 1) and (T/J = 1, 2, 3, 4, R3 = 0.01) has been given in Figures 11a,
Monte Carlo Simulation of Magnetic Properties … 65
11b and 11c, respectively with h/J = 0.2 and L = 4. The magnetization increase
with increasing the crystal field and decrease with increasing the size of
system such as that in Figure 11a for D < -0.4. The magnetization increases
with increasing the exchange interactions (see Figure 11b) for fixed size and
fixed crystal field. The magnetization decrease with increasing the
temperatures for fixed size of system and for fixed crystal field such as given
in Figure 11c. We have presented in Figure 12(a)-d, the total magnetization
versus the crystal field for mixed spins-5/2 and 2 Ising model on the Bethe
lattice with JS = 0.5, 1.5, 2.5, 3, T = 1, JS = -1, -2, -3, -3.5, T = 1, T = 1, 2.5,
3.5, 4.5, JS = 0.5 and JS = -0.5, respectively with Δ = 0.0 and h = 1. The total
magnetization increase with increasing of the absolute value of exchange
interactions JS( see Figures 12(a) and b) for -4 < D0. This behavior is
confirmed by the results given in Figure 12b. The total magnetization is
independent of exchange interaction JS for D > 0. The total magnetization
decrease with increasing of the temperature absolute value of exchange
interactions JS (see Figures 12c and 12d). Figure 13 illustrates, the total
magnetization versus the exchange interaction JS for T = 1, 2.5, 3.5, 4.5 with
Δ = 0.0 and h = 1. The total magnetization decrease with increasing of the
temperature for a fixed value of exchange interactions JS > 0. The inverse
behavior is observed for JS < 0. Finally, we have given in Figure 14, the
magnetic hysteresis cycle of kagomé lattice for T/J = 1, 5, 10, 20, Δ/J = 0 and
Δ/J = 0, -1.5, -2.5, -3.5 with R3 = 0.01 and L = 4. We have presented in Figure
15 (a,b)-(e,f), the variation of total magnetization and magnetization of each
superlattice with spin-5/2 and 2 versus the external magnetic field for D = 0, -
0.5, -1, respectively, with T = 1 and JS = 1. The increasing of absolute value
of the crystal field decreasing the magnetic coercive field for JS > 0. For
negative values of the crystal field such as Δ/J = −3.5 and the temperature T/J
= 20 again only one hysteresis loop and flatness exist in the system, seen in
Figures 14a and 14b, respectively. The coercive field decrease with increasing
the temperatures and crystal field values. The same behavior is observed in
previously works [44, 45]. This loop decreases with the increasing of the
temperature and at last, it disappears at a certain temperature value. Finally,
we have given in Figure 15(a,b)-(e,f), the variation of total magnetization and
magnetization of each superlattice with spin- and S versus the external
magnetic field for D = 0, -0.5, -1 with, temperature T = 1 and JS = +1. The
systems becomes superparamagnetic for D = -1, T = 1 and JS = +1 such as in
Figure 15e. The increase of absolute value of the crystal field decreasing the
magnetic coercive field for JS < 0. This behavior is similar to those obtained
66 R. Masrour and A. Jabar
by Ref. [46]. Figures 15e and 15e show that the coercive field is very low
in the ferromagnetic case and the system remain superparamagnitc rapidly
when compared with the case of antiferromagnetic case. The saturation
magnetization and remanent magnetization changing rapidly with the change
of sign of JS.
Figures 4a-d. The ground state phase diagrams of alternate mixed spin-5/2 and spin-2
Ising model on the Bethe lattice on the (Δ, J S), (h,JS) and (h, Δ), respectively for JS =
1.0 and (h, Δ) planes for JS = -1.0.
Monte Carlo Simulation of Magnetic Properties … 67
Figure 5. The thermal magnetization and magnetic susceptibility for JS = 1.0 (a), JS =
-1.0 (b) with Δ = 0.0 and h = 0.75.
Figure 6. (Continued).
68 R. Masrour and A. Jabar
Figure 6. The thermal total magnetization and magnetic susceptibility for JS = 1, 1.5, 2
(a), JS = -1, -1.5, -2 (b) with Δ = 0.0 and h = 0.75.
Figure 7. The thermal magnetization and magnetic susceptibility of kagomé lattice for
R3 = 0.1, 0.3, 0.5 with Δ/J = 0, h/J = 0 for L = 4.
Figure 8. The critical temperature versus the R3 of kagomé lattice for h/J = D/J = 0 and
with h/J = 0.
Monte Carlo Simulation of Magnetic Properties … 69
Figure 10. The magnetization versus the R3 of kagomé lattice for L = 1, 2, 3, 4, T/J = 5,
Δ/J = 0.0 (a), D/J = 0,-1, -1.2, -1.5, T/J = 5 (b) and T/J = 4, 5, 8, 10, Δ/J = 0 (c) with h/J
= 0 and L = 4.
Figure 11. The magnetization versus the crystal field D/J of kagomé lattice for L = 1, 2,
3, 4, T/J = 1, R3 = 0.01 (a), R3 = 0.01, 0.1, 0.2, 0.4, T/J = 1 (b) and T/J = 1, 2, 3, 4, R 3 =
0.01 (c) with h/J = 0.2 and L = 4.
Figure 12. The total magnetization versus the crystal field for JS = 0.5, 1.5, 2.5, 3, T =
1 (a), JS = -1, -2, -3, -3.5, T = 1(b), T = 1, 2.5, 3.5, 4.5, J S = 0.5 (c) and JS = -0.5 (d)
with Δ = 0.0 and h = 1.
Figure 13. The total magnetization versus the exchange interaction (JS) between spin-
5/2 and 2 for T = 1, 2.5, 3.5, 4.5 with Δ = 0.0 and h = 1.
Monte Carlo Simulation of Magnetic Properties … 73
Figure 14. The magnetic hysteresis cycle of kagomé lattice for T/J = 1, 5, 10, 20, Δ/J =
0 and Δ/J = 0, -1.5, -2.5, -3.5 with R3 = 0.01 and L = 4.
Figure 15. The variation of total magnetization and magnetization of each superlattice
with spin-5/2 and 2 vs the external magnetic field for D = 0 (a,b), -0.5 (c,d), -1 (e,f)
with temperature T = 1 and JS = -1.
Figure 16. The variation of total magnetization and magnetization of each superlattice
with spin-5/2 and 2 vs the external magnetic field for D = 0 (a,b), -0.5 (c,d), -1 (e,f)
with temperature T = 1 and JS = 1.
CONCLUSION
The magnetic properties of kagomé lattice with RKKY interactions and
mixed spins-5/2 and 2 Ising model on the Bethe lattice are investigated by
Monte Carlo simulation. The RKKY exchange interaction is obtained for
different distance between the two magnetic layers. The different phases are
found in the kagomé lattice with RKKY interactions and mixed spins-5/2 and
2 Ising model on the Bethe lattice. The critical temperature is obtained for
fixed size and increase with increasing the exchange interactions. The coercive
field decrease with increasing the temperatures and crystal filed. The effect of
FM and AFM in coercive field is studied. The coercive magnetic field
76 R. Masrour and A. Jabar
decrease with decreasing the crystal field for a positive and for a negative of
exchanges interactions in alternate and non alternate mixed spins. The
magnetic coercive field decreasing when the absolute value of the crystal field
increases for JS < 0 and for JS > 0. The field dependences of the hysteresis
curves show features characteristic of experimental studies [47–50].
Regions Phases
I (-2,-5/2)
II (+2,+5/2)
III (-2,+5/2)
IV (+2,-5/2)
V (-1,-5/2)
VI (+1,+5/2)
VII (0,-5/2)
VIII (0,+5/2)
IX (0,-3/2)
X (0,+3/2)
XI (0,-1/2)
XII (0,+1/2)
XIII (-1,+5/2)
XIV (+1,-5/2)
XV (-1,-5/2)
XVI (+1,+5/2)
XVII (-1,-1/2)
XVIII (+1,+1/2)
XIX (-2,-1/2)
XX (+2,+1/2)
XXI (-2,+1/2)
XXII (+2,-1/2)
XXIII (-2,-3/2)
XXIV (+2,+3/2)
XXV (-2,+3/2)
XXVI (+2,-3/2)
XXVII (-1,+1/2)
XXVIII (+1,-1/2)
XXIX (-1,+3/2)
XXX (+1,-3/2)
Monte Carlo Simulation of Magnetic Properties … 77
REFERENCES
[1] Y. Nakamura, Phys. Rev. B 62 (2000) 11742.
[2] Y. Nakamura, J. Phys.: Condens. Matter 12 (2000) 4067.
[3] P. Henelius, R. S. Fishman, Phys. Rev. B 78 (2008) 214405.
[4] Simon G. Carling, Peter Day, Polyhedron 20 (2001) 152.
[5] Yasen Hou, Qi Zhang, Yu Jia. Physica B 442 (2014) 52–56.
[6] Q. Zhang, G. Z. Wei, J. Magn. Magn. Mater. 253 (2002) 96.
[7] G. Z. Wei, Q. Zhang, Z. H. Xin, Y. Q. Liang, J. Magn. Magn. Mater.
277 (2004) 1.
[8] G. Z. Wei, Q. Zhang, J. Zhao, Y. W. Gu, Phys. B: Condens. Matter 381
(2006) 6.
[9] W. Jiang, C. Liu, Y. Jiang, Phys. A: Stat. Mech. Appl. 387 (2008) 4599.
[10] W. Jiang, W. Wang, F. Zhang, W. J. Ren, J. Appl. Phys. 105 (2009)
07E321.
[11] A. Yigit, E. Albayrak, Phys. Lett. A 372 (2008) 361.
[12] J. S. da Cruz Filho, M. Godoy, A. S. de Arruda, Phys. A: Stat. Mech.
Appl. 392 (2013) 6247.
[13] E. Albayrak, Phys. A: Stat. Mech. Appl. 375 (2007) 174.
[14] H. K. Mohamad, E. P. Domashevskaya, A. F. Klinskikh, Phys. A: Stat.
Mech. Appl. 388 (2009) 4713.
[15] J. Li, A. Du, G. Wei, Phys. Satus Solidi (b) 240 (2003) 610.
[16] J. Li, A. Du, G. Z. Wei, Phys. B: Condens. Matter 348 (2004) 79.
[17] K. L. Yao, J. W. Li, Z. L. Liu, H. H. Fu, L. Zu, Commun. Theor. Phys.
47 (2007) 741.
[18] A. Jabar, R. Masrour, A. Benyoussef, M. Hamedoun. J. Magn. Magn.
Mater. 397 (2016) 287–294.
[19] R. Masrour, A. Jabar, A. Benyoussef, M. Hamedoun. J. Magn. Magn.
Mater. 393 (2015) 151–156.
[20] N. De La Espriella, C. A. Mercado, J. C. Madera. J. Magn. Magn.
Mater. 401 (2016) 22–29.
[21] R. Masrour, A. Jabar, A. Benyoussef, M. Hamedoun. J. Magn. Magn.
Mater. 401 (2016) 700–705.
[22] R. Masrour, A. Jabar, A. Benyoussef, M. Hamedoun. J. Magn. Magn.
Mater. 401 (2016) 695–699.
[23] R. Masrour , A. Jabar, A. Benyoussef, M. Hamedoun. Chemical Physics
Letters 631–632 (2015) 92–96.
[24] Liqin Jiang, Jianmin Zhang, Zhigao Chen, Qian Feng, Zhigao Huang.
Physica B 405 (2010) 420–424.
78 R. Masrour and A. Jabar
Buchen.
Published/Created Boca Raton, FL: CRC Press, c2012.
Description xvii, 278 p.: ill.; 25 cm.
ISBN 9781420091007 (hardback)
142009100X (hardback)
LC classification HG6024.A3 B83 2012
Summary "In an easy-to-understand, nontechnical yet
mathematically elegant manner, An Introduction to
Exotic Option Pricing shows how to price exotic
options, including complex ones, without performing
complicated integrations or formally solving partial
differential equations (PDEs). The author incorporates
much of his own unpublished work, including ideas
and techniques new to the general quantitative finance
community.The first part of the text presents the
necessary financial, mathematical, and statistical
background, covering both standard and specialized
topics. Using no-arbitrage concepts, the Black-Scholes
model, and the fundamental theorem of asset pricing,
the author develops such specialized methods as the
principle of static replication, the Gaussian shift
theorem, and the method of images. A key feature is
the application of the Gaussian shift theorem and its
multivariate extension to price exotic options without
needing a single integration.The second part focuses
on applications to exotic option pricing, including
dual-expiry, multi-asset rainbow, barrier, lookback,
and Asian options. Pushing Black-Scholes option
pricing to its limits, the author introduces a powerful
formula for pricing a class of multi-asset, multiperiod
derivatives. He gives full details of the calculations
involved in pricing all of the exotic options.Taking an
applied mathematics approach, this book illustrates
how to use straightforward techniques to price a wide
range of exotic options within the Black-Scholes
framework. These methods can even be used as control
variates in a Monte Carlo simulation of a stochastic
volatility model"--Provided by publisher.
"Preface This book is a collection of a large amount of
82 Bibliography
Excel simulations
LCCN 2013938520
Type of material Book
Personal name Verschuuren, G. M. N. (Geert M. N.), author.
Main title Excel simulations / Dr. Gerald M. Verschuuren.
Published/Produced Uniontown, OH: Holy Macro! Books, [2014]
Description iv, 172 pages: illustrations; 27 cm
Links Contributor biographical information http://www.loc.
gov/catdir/enhancements/fy1511/2013938520-b.html
Publisher description http://www.loc.gov/catdir/
enhancements/fy1511/2013938520-d.html
ISBN 9781615470228
1615470220
LC classification HF5548.4.M523 V47 2014
Summary "Covering a variety of Excel simulations, from
gambling to genetics, this introduction is for people
interested in modeling future events, without the cost
of an expensive textbook. The simulations covered
offer a fun alternative to the usual Excel topics and
include situations such as roulette, password cracking,
sex determination, population growth, and traffic
patterns, among many others"--"--Amazon.com,
viewed December 31, 2013.
Contents Gambling -- Statistics -- Genetics -- Financial --
Expansion -- Monte Carlo simulation -- Iterations --
Extras.
Subjects Microsoft Excel (Computer file)
Digital computer simulation.
102 Bibliography
Computer simulation.
Notes "Using Excel to model risk, investments, genetics,
growth, gambling and Monte Carlo analysis"--Cover.
Includes index.
[2012]
Description xix, 314 pages; 24 cm.
Links Cover image http://catalogimages.wiley.com/
images/db/jimages/9781118175446.jpg
ISBN 9781118175446 (pbk.)
LC classification HG106 .C485 2012
Cover title Financial modeling with Oracle Crystal Ball and Excel
+ website
Summary "Updated look at financial modeling and Monte Carlo
simulation with software by Oracle Crystal Ball. This
revised and updated edition of the bestselling book on
financial modeling provides the tools and techniques
needed to perform spreadsheet simulation. It answers
the essential question of why risk analysis is vital to
the decision-making process, for any problem posed in
finance and investment. This reliable resource reviews
the basics and covers how to define and refine
probability distributions in financial modeling, and
explores the concepts driving the simulation modeling
process. It also discusses simulation controls and
analysis of simulation results. The second edition of
Financial Modeling with Crystal Ball and Excel
contains instructions, theory, and practical example
models to help apply risk analysis to such areas as
derivative pricing, cost estimation, portfolio allocation
and optimization, credit risk, and cash flow analysis. It
includes the resources needed to develop essential
skills in the areas of valuation, pricing, hedging,
trading, risk management, project evaluation, credit
risk, and portfolio management. Offers an updated
edition of the bestselling book covering the newest
version of Oracle Crystal Ball Contains valuable
insights on Monte Carlo simulation--an essential skill
applied by many corporate finance and investment
professionals Written by John Charnes, the former
finance department chair at the University of Kansas
and senior vice president of global portfolio strategies
at Bank of America, who is currently President and
Chief Data Scientist at Syntelli Solutions, Inc. Risk
104 Bibliography
LCCN 2011050888
Type of material Book
Personal name Peterson, Steven P.
Main title Investment theory and risk management / Steven P.
Peterson.
Published/Created Hoboken, N.J.: Wiley, 2012.
Description xix, 441 p.: ill.; 24 cm.
Links Cover image http://catalogimages.wiley.com/
images/db/jimages/9781118129593.jpg
ISBN 9781118129593 (hardback)
1118129598 (hardback)
LC classification HG4529 .P478 2012
Summary "A unique perspective on applied investment theory
and risk management from the Senior Risk Officer of a
major pension fundInvestment Theory and Risk
Management is a practical guide to today's investment
environment. The book's sophisticated quantitative
methods are examined by an author who uses these
methods at the Virginia Retirement System and
teaches them at the Virginia Commonwealth
University. In addition to showing how investment
performance can be evaluated, using Jensen's Alpha,
Sharpe's Ratio, and DDM, he delves into four types of
optimal portfolios (one that is fully invested, one with
targeted returns, another with no short sales, and one
with capped investment allocations). In addition, the
book provides valuable insights on risk, and topics
such as anomalies, factor models, and active portfolio
management. Other chapters focus on private equity,
structured credit, optimal rebalancing, data problems,
and Monte Carlo simulation. Contains investment
theory and risk management spreadsheet models based
on the author's own real-world experience with stock,
bonds, and alternative assets Offers a down-to-earth
guide that can be used on a daily basis for making
common financial decisions with a new level of
quantitative sophistication and rigor Written by the
Director of Research and Senior Risk Officer for the
Virginia Retirement System and an Associate
Bibliography 109
Numerical analysis
LCCN 2011014232
Type of material Book
Personal name Sauer, Tim.
Main title Numerical analysis / Timothy Sauer.
Edition 2nd ed.
Published/Created Boston: Pearson, c2012.
Description xvi, 646 p.: ill.; 26 cm.
ISBN 9780321783677
0321783670
LC classification QA297 .S348 2012
Contents Machine generated contents note: ch. 0 Fundamentals -
- 0.1.Evaluating a Polynomial -- 0.2.Binary Numbers -
- 0.2.1.Decimal to binary -- 0.2.2.Binary to decimal --
0.3.Floating Point Representation of Real Numbers --
0.3.1.Floating point formats -- 0.3.2.Machine
representation -- 0.3.3.Addition of floating point
numbers -- 0.4.Loss of Significance -- 0.5.Review of
Calculus -- Software and Further Reading -- ch. 1
Solving Equations -- 1.1.The Bisection Method --
1.1.1.Bracketing a root -- 1.1.2.How accurate and how
fast? -- 1.2.Fixed-Point Iteration -- 1.2.1.Fixed points
of a function -- 1.2.2.Geometry of Fixed-Point
Iteration -- 1.2.3.Linear convergence of Fixed-Point
Iteration -- 1.2.4.Stopping criteria -- 1.3.Limits of
Accuracy -- 1.3.1.Forward and backward error --
1.3.2.The Wilkinson polynomial -- 1.3.3.Sensitivity of
root-finding -- 1.4.Newton's Method -- 1.4.1.Quadratic
convergence of Newton's Method -- 1.4.2.Linear
convergence of Newton's Method -- 1.5.Root-Finding
without Derivatives -- 1.5.1.Secant Method and
variants -- 1.5.2.Brent's Method -- Reality Check 1
Kinematics of the Stewart platform -- Software and
Further Reading -- ch. 2 Systems of Equations --
2.1.Gaussian Elimination -- 2.1.1.Naive Gaussian
elimination -- 2.1.2.Operation counts -- 2.2.The LU
Factorization -- 2.2.1.Matrix form of Gaussian
elimination -- 2.2.2.Back substitution with the LU
factorization -- 2.2.3.Complexity of the LU
Bibliography 121
The Monte Carlo simulation method for system reliability and risk
analysis.
LCCN 2012950369
Type of material Book
Main title The Monte Carlo simulation method for system
reliability and risk analysis.
Published/Created New York: Springer, 2012.
Links Contributor biographical information http://www.loc.
gov/catdir/enhancements/fy1316/2012950369-b.html
Publisher description http://www.loc.gov/catdir/
enhancements/fy1316/2012950369-d.html
Table of contents only http://www.loc.gov/catdir/
enhancements/fy1316/2012950369-t.html
ISBN 9781447145875
by publisher.
"This is a scholarly double-blind peer-reviewed edited
book. Quality was designed-in through the team
selection and review processes. Over 40 experienced
practitioner-scholars and subject matter experts were
selected from accredited universities and respected
organizations around the world. For example, our
foreword author Dr. Joseph F. Hair Jr. is a well-known
research methods textbook author. A new state-of-the-
art research design typology model was developed to
guide authors and readers. The authors used visual
techniques - tables and diagrams - to appease all
learning styles and especially English Second
Language Learners. Additionally, the authors
integrated and applied authentic examples into their
chapters based on their experience or a published
scholarly study. A visual chapter map in each chapter
shows where the design theory is positioned on the
research ideology continuum and which research
design typology components are addressed. The
closing chapter on gaps in future practices contains an
interesting surprise"-- Provided by publisher.
Contents Machine generated contents note: -- Preliminary
Material -- i. Foreword: Vision and Reality; Joseph F.
Hair, Jr -- ii. Preface: What's Unique for Practitioners;
Kenneth D. Strang -- Research Design Guidelines -- 1.
Why Practitioner-Scholars Need a Research Design
Typology; Kenneth D. Strang -- 2. Articulating a
Research Design Ideology; Kenneth D. Strang -- 3.
Developing a Goal-Driven Research Strategy; Kenneth
D. Strang -- 4. Matching Research Method with
Ideology and Strategy; Kenneth D. Strang -- 5.
Selecting Research Techniques for a Method and
Strategy; Kenneth D. Strang -- 6. Design Issues in
Cross-Cultural Research: Suggestions for Researchers;
Linda Brennan, Lukus Parker, Dang Nguyen, Torgeir
Aleti -- 7. Establishing Rationale and Significance of
Research; Judith Hahn -- 8. Organizing and
Conducting Scholarly Literature Reviews; Linnaya
142 Bibliography
Understanding probability
LCCN 2012010536
Type of material Book
Personal name Tijms, H. C.
Main title Understanding probability / Henk Tijms.
Edition 3rd ed.
Published/Created New York: Cambridge University Press, 2012.
Description x, 562 p.: ill.; 23 cm.
Bibliography 145
Subjects Probabilities.
Mathematical analysis.
Chance.
Mathematics / Probability & Statistics / General
Notes Includes bibliographical references (p. 556-557) and
index.
Will you be alive 10 years from now?: and numerous other curious
questions in probability: a collection of not so well-known mathematical
mind-benders (with solutions, with one exception)
LCCN 2013010221
Type of material Book
Personal name Nahin, Paul J.
Main title Will you be alive 10 years from now?: and numerous
other curious questions in probability: a collection of
not so well-known mathematical mind-benders (with
solutions, with one exception) / Paul J. Nahin.
Published/Produced Princeton: Princeton University Press, [2014]
Description xxvi, 220 pages; 25 cm
ISBN 9780691156804 (hardback)
LC classification QA273.25 .N344 2014
Summary "What are the chances of a game-show contestant
finding a chicken in a box? Is the Hanukkah dreidel a
fair game? Will you be alive ten years from now?
These are just some of the one-of-a-kind probability
puzzles that acclaimed popular math writer Paul Nahin
offers in this lively and informative book.Nahin brings
probability to life with colorful and amusing historical
anecdotes as well as an electrifying approach to
solving puzzles that illustrates many of the techniques
that mathematicians and scientists use to grapple with
probability. He looks at classic puzzles from the past--
from Galileo's dice-tossing problem to a disarming
dice puzzle that would have astonished even Newton--
and also includes a dozen challenge problems for you
to tackle yourself, with complete solutions provided in
the back of the book.Nahin then presents twenty-five
unusual probability puzzlers that you aren't likely to
find anywhere else, and which range in difficulty from
Bibliography 147
*
The full version of this chapter can be found in New Developments in Photon and Materials
Research, edited by Joon I. Jang, published by Nova Science Publishers, Inc, New York, 2013.
†
E-mail address: james.chow@rmp.uhn.on.ca.
150 Related Nova Publication
investigations in the bone and air inhomogeneity when photon beams in the
kilovoltage (kV) and megavoltage (MV) range were used in radiotherapy. For
the kV photon beams, the variations of the bone and mucosal doses with
different soft tissue and bone thicknesses mimicking the oral or nasal cavity
were examined using Monte Carlo simulation. Inhomogeneous phantoms
containing water, bone and air layers were irradiated with photon beams. It is
found that the 105 kVp photon beam produced more than twice the bone dose
than the 220 kVp, when the water thickness at the phantom surface was small
(0.5 cm). In another work, the dosimetric impact of surface dose reduction due
to the loss of backscatter from the bone interface was evaluated using the 105
kVp photon beam. It is concluded that the decrease of surface dose results in
an overestimation of prescribed dose at the patient’s surface, and might be a
concern when using superficial photon beam to treat skin tumours in sites such
as forehead, chest wall and kneecap. On the other hand, the variations of the
dose and dose distribution, due to the photon beam energy and presence of
inhomogeneity in a small-animal irradiation, were studied using photon beam
energies of 50, 100, 150, 200, 250, 300 and 1250 kV. The results showed that
the increase of mean dose due to the bone dose enhancement could be a
dosimetric concern, when using kV photon beam in the small-animal
irradiation and model. For the MV photon beams, the dependences of mucosal
dose in the oral or nasal cavity on the beam energy, beam angle and mucosal
thickness were studied. Two inhomogeneous phantoms namely, slab and
cylindrical, were used with the 6 and 18 MV photon beams. Results from the
slab phantom showed that the mucosal dose with bone varied with the beam
energy, beam angle and mucosal thickness. For the more realistic cylindrical
phantom, it is found that the mucosal surface doses decreased slightly with an
increase of mucosal thickness, when the beam angle was equal to zero. These
dosimetric variations are important to consider when improving the treatment
strategy to minimize the mucosal complications in head-and-neck
radiotherapy.
INDEX
biosensors, 129
A blueprint, 140
bonds, 108
access, 10, 13, 125
bone, 149
acid, 114, 128, 129, 144
boundary value problem, 123
AFM, 57, 63, 75
Brownian motion, 123, 145
algorithm, 2, 5, 6, 9, 14, 15, 17, 22, 23, 26,
business environment, 84
59, 105, 124, 149
businesses, 13
angiosperm, 129
anisotropy, 129
ANOVA, 86 C
antiviral drugs, 52
applied mathematics, 81 C++, 14, 91
arbitrage, 81, 82 calculus, 115, 145
architect, 140 calibration, 89, 90
arithmetic, 16, 17, 22, 26, 27, 123 call centers, 135
Artificial Neural Networks, 5 cancerous cells, 149
assessment, vii, 1, 3, 33, 34, 37, 92, 98, 100, case study, 3, 26, 128
132 cash flow, 103, 127
assets, 108, 112, 125, 133, 137, 145 classification, 9, 32, 79, 81, 83, 84, 87, 90,
atoms, 58, 61, 143 91, 96, 99, 100, 101, 102, 103, 104, 105,
107, 108, 112, 113, 114, 117, 118, 120,
125, 126, 127, 128, 130, 131, 132, 133,
B 134, 136, 140, 143, 144, 145, 146
clinical trials, 44, 49
baby boomers, 136
Cloud computing, 13
bacterial pathogens, 129
clustering, 80, 125, 131
base, 93, 99, 139
coefficient of variation, 6
Bayesian methods, 83, 87
communication, 9, 14, 22, 24, 115
beams, 144, 150
compensation, 56, 63
bioinformatics, 116
compilation, 117, 134
biomaterials, 51
152 Index
heterogeneity, 52
F hexagonal lattice, 56
high performance computing, 5
Fermi level, 58
Hill equation, viii, 36, 41
ferromagnetic, 56, 57, 59, 63
HIV/AIDS, viii, ix, 35, 36, 37, 38, 39, 40,
ferromagnetism, 64
42, 43, 46, 47, 49, 51, 52, 53, 54
FFT, 123
HIV-1, 37, 51, 52, 53
field theory, 56, 63, 143
human, 36, 48
financial, 80, 81, 83, 90, 103, 104, 105, 108,
human immunodeficiency virus, 36
112, 119, 135, 139
hybrid, 14, 26, 27
financial crisis, 112
hypothesis test, 125
financial markets, 83, 112
hysteresis, ix, 56, 65, 73, 76
finite element method, 133
flatness, 65
fluctuations, 43, 44, 47, 49 I
fluorescence, 129
forecasting, 32, 33 identification, 126
formula, 81, 121 ideology, 141
foundations, 114 image, 81, 83, 90, 96, 99, 103, 108, 112,
Fourier transform technique, 91 117, 125, 133, 136, 145
fragility, 136 immune system, 48, 52
freedom, 134 immunodeficiency, 36, 49
in vitro, viii, 36, 37, 40, 48
in vivo, viii, 36, 37, 53
G income, 136, 137
independence, 20
Galileo, 146
individuals, viii, 36
gambling, 101, 102
inertial navigation system, 106
game theory, 116
infection, viii, ix, 35, 36, 37, 38, 41, 42, 47,
gasification, 132
49, 50, 51, 52, 53
Gaussian random variables, 82
information processing, 9
genetics, 101, 102
infrastructure, 13
genome, 139
inhibition, viii, 36, 41, 47, 50
genomics, 80
inhomogeneity, 149
geometry, 117
initial state, 6, 16
gold nanoparticles, vii, viii, 35, 37, 38, 50
initiation, 43
GPS, 106, 107, 122
institutions, 90
graduate education, 115
integrated circuits, 128
graduate students, 126
integration, 81, 82, 84, 93, 122, 125
grid computing, 13
interface, 2, 13, 21, 23, 150
growth, viii, 36, 101, 102
internalization, 44
interprocessor communication, 9
H investment, 102, 103, 108, 112, 132, 137
investors, 112
Hamiltonian, 57, 58, 59 ion channels, 129
hedging, 83, 89, 90, 103 irradiation, 150
154 Index