You are on page 1of 23

Dr.

Sitanshu Sekhar Sahu


Dept. of Electronics & Communication Engineering
Birla Institute of Technology, Mesra, Ranchi
Email: sssahu@bitmesra.ac.in
Teaching Learning Based Optimization
(TLBO) R.V. Rao, 2011

 Based on Philosophy of Teaching and Learning: Knowledge


transfer in a classroom environment
Teaching Learning Based Optimization
(TLBO)
 It is based on the effect of the influence of a teacher on the output of
the learners in a class. The output of class is considered in terms of
results or grades.
 The teacher is normally considered as a highly learned person who
shares his or her knowledge with the learners. The quality of a
teacher’s capability affects the outcome of learners.
 Learners also learn from interaction between themselves.

The algorithm mimics the teaching-learning ability of the


teacher and learners in a classroom.

3
Teaching Learning Based Optimization
(TLBO)
 It is a population based metaheuristic optimization technique.

 In this algorithm, the population is considered as a group of learner


and the whole learning process is divided into two part such as
‘Teacher Phase’ and ‘Learner Phase’.

 The most knowledgeable person in the society is considered as


teacher and in every iteration the best learner is considered as a
teacher.
 In ‘Teacher Phase’ the learner group learns from teacher.

 In ‘Learner Phase’ they learn by having interaction with each others.

4
Teaching Learning Based Optimization
(TLBO)
 The teacher tries to enhance the knowledge of the entire learner
in accordance with his or her capability.

 The transfer of knowledge also depends on the capability of the


class. In other word the learner will gain knowledge according to
the quality of teaching delivered and also on their quality.

 Different designed variables of the optimization problem are


analogous to different subjects offered to the students (learners).
The results of each learner are equivalent to the fitness of the
problem.
Teaching Learning Based Optimization
(TLBO)
 Teacher phase:
 New solution is generated using the best one the class and mean of
the population.
 Greedy Selection: Accept the new solution if better than the
current solution.

 Learner Phase:
 New solution is generated using a partner solution (randomly
selected solutions from the class)
 Greedy selection

Each solution undergoes Teacher phase followed by


Learner phase.
Mathematical Model of TLBO

Teachers phase
𝒙𝒏𝒆𝒘 = 𝒙 + 𝒓 ∗ (𝒙𝑻𝒆𝒂𝒄𝒉𝒆𝒓 − 𝑻𝑭 ∗ 𝒙𝑴𝒆𝒂𝒏 )

TF=round [1+rand(0,1)]
TF: Teaching factor (TF) ; its value can be either 1 or 2 and it is decided
randomly with equal probability

Learners phase
𝒙𝒊 + 𝒓𝟏 ∗ 𝒙𝒊 − 𝒙𝒋 𝒊𝒇 𝒇 𝒙𝒊 < 𝒇(𝒙𝒋 )
𝒙𝒏𝒆𝒘 =
𝒙𝒋 + 𝒓𝟐 ∗ 𝒙𝒋 − 𝒙𝒊 𝒐𝒕𝒉𝒆𝒓𝒘𝒊𝒔𝒆
Flowchart of TLBO

8
Himmelblau Function
Minimize

Solution:
𝑓 𝑥 =0
At 𝑥1 =2, 𝑥2 = 3

9
TLBO
Initial Population: 5 (number of learners)
Learner 𝒙𝟏 𝒙𝟐 f(x)
1 3.22 0.403 13.139
2 0.191 2.289 77.710
3 3.182 0.335 14.024
4 1.66 4.593 261.57
5 2.214 0.867 43.641
Mean 2.093 1.697

Teacher
(lowest fitness value)
10
Teacher Phase
Assume 𝒓𝟏 = 𝟎. 𝟐𝟓 𝒇𝒐𝒓 𝒙𝟏 and 𝒓𝟐 = 𝟎. 𝟒𝟑 𝒇𝒐𝒓 𝒙𝟐 𝑻𝒇 = 1
Updated students and the Fitness function
Learner 𝒙𝟏 𝒙𝟐 f(x)
1 3.501 -0.153 13.303
2 0.472 1.732 94.221
3 3.463 -0.221 12.765
4 1.941 4.036 136.437
5 2.495 0.310 39.326
Learner 2 (𝒙𝟏 )= 0.191+0.25 (3.22-1*2.093)=0.472
Learner 2 (𝒙𝟐 )=2.289+0.43(0.403-1*1.697)=1.732
𝒙𝒏𝒆𝒘 = 𝒙 + 𝒓 ∗ (𝒙𝑻𝒆𝒂𝒄𝒉𝒆𝒓 − 𝑻𝑭 ∗ 𝒙𝑴𝒆𝒂𝒏 )
TF=round [1+rand(0,1)]
Teacher Phase
Updated students and the Fitness function
Learner 𝒙𝟏 𝒙𝟐 f(x)
1 3.22 0.403 13.139
2 0.191 2.289 77.710
3 3.463 -0.221 12.765
4 1.941 4.036 136.437
5 2.495 0.310 39.326

Best values of the fitness function are considered for Learners phase.
Learner Phase
Any student can interact with any other student
for knowledge transfer and interaction can be
done in a random manner 𝒓𝟏 = 0.47 𝒓𝟐 =0.33
Learner 𝒙𝟏 𝒙𝟐 f(x) Learners
Interaction
1 3.22 0.403 13.139 1&2
2 0.191 2.289 77.710 2&5
3 3.463 -0.221 12.765 3&1
4 1.941 4.036 136.437 4&5
5 2.495 0.310 39.326 5&4

𝒙𝒊 + 𝒓𝟏 ∗ 𝒙𝒊 − 𝒙𝒋 𝒊𝒇 𝒇 𝒙𝒊 < 𝒇(𝒙𝒋 )
𝒙𝒏𝒆𝒘 =
13
𝒙𝒋 + 𝒓𝟐 ∗ 𝒙𝒋 − 𝒙𝒊 𝒐𝒕𝒉𝒆𝒓𝒘𝒊𝒔𝒆
Learner Phase New fitness value

Updated students and the Fitness function


𝒙𝟏 𝒙𝟐 f(x) Learners
Interaction
4.643 -0.219 112.324 1&2
1.274 1.636 69.206 2&5
3.578 -0.426 12.391 3&1
2.202 2.807 20.674 4&5
2.756 -0.919 30.241 5&4
Learner 1 (𝒙𝟏 )= 3.22+0.47 (3.22-0.191)=4.643 𝒓𝟏 = 0.47 𝒓𝟐 =0.33
Learner 1 (𝒙𝟐 )=0.403+0.33(0.403-2.289)=-0.2193

Learner 2 (𝒙𝟏 )= 0.191+0.47 (2.495-0.191)=1.2742


14Learner 2 (𝒙𝟐 )=2.289+0.33(0.310-2.289)=1.636
Learner phase
Updated students and the Fitness function
𝒙𝟏 𝒙𝟐 f(x)
3.22 0.403 13.139
1.27 1.636 69.206
3.57 -0.426 12.391
2.20 2.807 20.674
2.75 -0.919 30.241

Teacher for 2nd iteration

Best learner value and best of the fitness function between teachers
15phase and leaners phase are considered for 2nd iteration.
Updated Students and fitness function
after 1st Iteration
Learner 𝒙𝟏 𝒙𝟐 f(x)
1 3.22 0.403 13.139
2 0.191 2.289 77.710
3 3.182 0.335 14.024
4 1.66 4.593 261.57
5 2.214 0.867 43.641
Learner 𝒙𝟏 𝒙𝟐 f(x)
1 3.22 0.403 13.139
2 1.27 1.636 69.206
3 3.57 -0.426 12.391
4 2.20 2.807 20.674
16
5 2.75 -0.919 30.241
Issues with TLBO
 The variables need to be fit within the bound in both the teacher phase
and learner phase

 Need to remove the duplicate solutions/ same fitness values


two solutions with identical set of variables
two solutions different set of variables but same fitness value

 Multiple solutions may have same partner in TLBO

 Number of fitness function evaluation if duplicate removal is involved.

17
Modified TLBO
Teaching-learning process is an iterative process wherein the continuous interaction takes
place for the transfer of knowledge.
Every time a teacher interacts with a student he/she finds that the student is able to recall
part of the lessons learnt from the last session. This is mainly due to the physiological
phenomena of neurons in the brain.
A new parameter known as “weight” is included in the original TLBO
𝒘𝒎𝒂𝒙 −𝒘𝒎𝒊𝒏
𝒘 = 𝒘𝒎𝒂𝒙 -( )*g
𝑮
Where 𝒘𝒎𝒂𝒙 , 𝒘𝒎𝒊𝒏 are the maximum and minimum values of weight factor w [0.9,
0.1], g is the current iteration number and G is the maximum number of allowable
iterations.
𝒙𝒏𝒆𝒘 = 𝒘 ∗ 𝒙 + 𝒓 ∗ (𝒙𝑻𝒆𝒂𝒄𝒉𝒆𝒓 − 𝑻𝑭 ∗ 𝒙𝑴𝒆𝒂𝒏 )

TF=round [1+rand(0,1)]
𝒘 ∗ 𝒙𝒊 + 𝒓𝟏 ∗ 𝒙𝒊 − 𝒙𝒋 𝒊𝒇 𝒇 𝒙𝒊 < 𝒇(𝒙𝒋 )
𝒙𝒏𝒆𝒘 =
𝒘 ∗ 𝒙𝒋 + 𝒓𝟐 ∗ 𝒙𝒋 − 𝒙𝒊 𝒐𝒕𝒉𝒆𝒓𝒘𝒊𝒔𝒆
18
Modified TLBO
Merit:
• It is reliable, accurate and robust.
• Total computational time is less.
• Consistency is high.
• It is quite efficient as comparison to basic TLBO .
• MTLBO finds better or equal solutions much faster than other variants.
• It gives better performance with less computational time for the
problems with high dimensions.

Demerit:
It converges quickly if proper precaution is not taken.

19
Elitist TLBO
 During every generation the worst solutions are replaced by the elite
(best) solutions.

 In TLBO, after replacing the worst solutions with elite solutions at the
end of learner phase, if the duplicate solutions exist then it is necessary
to modify the duplicate solutions in order to avoid trapping in the local
optima.

 The, duplicate solutions are modified by mutation on randomly selected


dimensions of the duplicate solutions before executing the next
generation.

20
24 July 2011
Resource
 R.V. Rao, V.J. Savsani, J. Balic, “Teaching–learning-based optimization algorithm for
unconstrained and constrained real-parameter optimization problems”, Engineering
Optimization (Taylor & Francis), Volume 44, Issue 12, 2012, 1447-1462

 R.V. Rao, V. Patel, “An elitist teaching-learning-based optimization algorithm for solving complex
constrained optimization problems” International Journal of Industrial Engineering
Computations, Volume 3, Issue 4, 2012, 535-560.

 R.V. Rao, V.D. Kalyankar , “Parameter optimization of machining processes using a new
optimization method” Materials and Manufacturing Processes, Volume 27, Issue 9,
2012, 978-985
 R.V. Rao, “ Review of applications of TLBO algorithm and a tutorial for beginners to solve
the unconstrained and constrained optimization problems”. Decision Science Letters, 5(1),
1-30, 2016.

Book: Teaching-Learning-Based Optimization (TLBO) Algorithm And Its


Engineering Applications. Springer International Publishing,
Switzerland, 2016, DOI:10.1007/978-3-319-22732-0.

https://sites.google.com/site/tlborao/

You might also like