76

CHAPTER 4

MATHEMATICAL MODELING AND ANALYSIS

In this chapter a mathematical model is developed to have a deeper

understanding of the relationship between the POME ink density and color
values which determine the print quality. A mathematical regression model
has been applied to calculate ink density 6D and color density variation
6Eab. The regression model has been assessed using hypothetical tests of R2,
T-value and F-value ANOVA to ascertain the fitness of the regression model.

4.1 MATHEMATICAL MODELING USING ORIGINPRO8

4.1.1 Curve Fitting

Curve fitting is the process of constructing a curve, or mathematical

function that has the best fit to a series of data points, possibly subject to
constraints. Curve fitting can involve either interpolation, where an exact fit
to the data is required, or smoothing, in which a "smooth" function is
constructed that approximately fits the data.

4.1.2 Regression Analysis

Regression analysis focuses more on questions of statistical

inference such as how much uncertainty is present in a curve that is fit to data
observed with random errors. Fitted curves can be used as an aid for data
visualization, to infer values of a function where no data are available, and to
summarize the relationships among two or more variables. Extrapolation
 77

refers to the use of a fitted curve beyond the range of the observed data, and is
subject to a degree of uncertainty since it may reflect the method used to
construct the curve as much as it reflects the observed data.

4.1.3 Different Types of Curve Fitting

Fitting functions to data points

Most commonly, one fits a function of the form y=f(x).

Fitting lines and polynomial functions to data points

Starting with a first degree polynomial equation:

y = ax + b

This is a line with slope a. A line will connect any two points, so a
first degree polynomial equation is an exact fit through any two points with
distinct x coordinates.

If the order of the equation is increased to a second degree

polynomial, the following results:

Y = ax2 + bx + c

This will exactly fit a simple curve to three points.

If the order of the equation is increased to a third degree

polynomial, the following is obtained:

Y = ax3 + bx2 + cx + d

This will exactly fit four points.

 78

Figure 4.1 Different levels of polynomial

Figure 4.1 shows that a first degree polynomial equation could also
be an exact fit for a single point and an angle while the third degree
polynomial equation could also be an exact fit for two points, an angle
constraint, and a curvature constraint. Many other combinations of constraints
are possible for these and for higher order polynomial equations.

4.1.4 Introduction to Origin & OriginPro

Origin is an easy-to-use software application that provides powerful

data analysis and publication-quality graphing capabilities tailored to the
needs of scientists and engineers. OriginPro offers all of the features of Origin
plus extended analysis tools for Peak Fitting, Surface Fitting, Statistics, Signal
Processing, and Image Handling OriginPro is the software of choice for those
who want a single data analysis and graphing solution.

Origin is a proprietary computer program for interactive scientific

graphing and data analysis. It is produced by OriginLab Corporation, and runs
on Microsoft Windows. It has inspired several platformindependent
 79

opensource clones like QtiPlot or SciDAVis. Graphing support in Origin

includes various 2D/3D plot types.

Data analyses in Origin include statistics, signal processing, curve

fitting and peak analysis. Origin's curve fitting is performed by the nonlinear
least squares fitter which is based on the Levenberg–Marquardt algorithm.

Origin imports data files in various formats such as ASCII text,

Excel, NI TDM, DIADem, NetCDF, SPC, etc. It also exports the graph to
various image file formats such as JPEG, GIF, EPS, TIFF, etc. There is also a
built-in query tool for accessing database data via ADO.

4.2 STEPS FOR MATHEMATICAL MODELING (REGRESSION

ANALYSIS) IN ORIGINPRO

A new OrignPro project has to be created using the menu options

File -> New -> Project. Upon the creation of the new project, the following
window “Book1” will appear as shown in Figure 4.2.

Figure 4.2 OriginPro New Window

 80

Type or Paste the data of experimented results of 6D and 6Eab

obtained from Xrite spectrometer for gloss coated paper for CMYK colors as
shown in the following Figure 4.3:

Figure 4.3 OriginPro Gloss Coated Paper Data

To create the mathematical models using curve fitting, we have

chosen the menu options Analysis -> Fitting -> Nonlinear Curve Fit -
>Opened Dialog as shown in the following Figure 4.4:

Figure 4.4 OriginProNonlinear Curve Fit

 81

The following dialog appeared as shown in Figure 4.5, where the

details of the data and type of modeling required would be given:

Figure 4.5 OriginPro Modeling Selection

For function selection, we have chosen the menu options Function

Selection -> Category: Polynomial and Function: Line as shown in the
following Figure 4.6:

Figure 4.6 OriginProPolynomial and Function Selection

 82

For selecting the data, we have chosen the menu options Data
Selection -> Input Data as shown in the following Figure 4.7:

Figure 4.7 OriginPro Menu Option Selection

We have selected the data using the options Select: [Book 1]Sheet1
and Click Add button as shown in the following Figure 4.8:

Figure 4.8 OriginProPolynomial Data Addition

 83

By clicking the ‘Ok’ button, the required data would be selected for
the regression analysis. By clicking the ‘Fit’ button as shown in the following
Figure 4.9, the regression analysis would be started.

Figure 4.9 OriginPro Regression Analysis Selection

On completion of the regression analysis, the dialog as shown in

the following Figure 4.10would appear. We have clicked the ‘Ok’ button for
viewing the parameters for the mathematical models and the fitted curve plot.

Figure 4.10 OriginPro Regression Analysis Curve Fitting

 84

We could find the parameter values A, B, R2 and F as shown in the

following Figure 4.11:

Figure 4.11 OriginProSelection R2and F Values

By scrolling down the window, we could find the fitted curves plot
as shown in the following Figure 4.12:

Figure 4.12 OriginPro Fitted Plot Curve

 85

By double-clicking the curve window, we could get the enlarged

curve as shown in the following Figure 4.13:

Figure 4.13 OriginPro Enlarged Fitted Plot Curve

Figure 4.14 shows the complete data set of 100 data was generated
after curve fitting Regression Analysis by the OriginPro software from the
given data of 15 data available from spectrometer reading in “FitNLCurve1”.

Figure 4.14 OriginProManuplated 100 Data

 86

For analyzing the goodness of fit of the generated mathematical

model using T-statistics, the hypothesis testing was done by choosing the
menu Statistics->Hypothesis Testing->Pair-Sample-Test->Open Dialog as
shown in the following Figure 4.15:

Figure 4.15 OriginProSlection of t-Test

Figure 4.16 shows the following dialog box to proceed further to

analysis of regression model for 100 data created by the OrginPro software.

Figure 4.16 OriginProPolynomial Data Addition

 87

Figure 4.17 shows the given Input data as 1st Data Range: B(Y) and
2nd Data Range: A(X) and was clicked ‘Ok’ button to find the T-statistic
values.

Figure 4.17 OriginPro Input Data of B(Y) and A(X)

The value of “T-statistic” was available as shown in the following

Figure 4.18.

Figure 4.18 OriginPro t-Statistic Values

 88

Parabola curve fitting

For the selection of parabola curve, we have chosen the menu

options Function Selection -> Category: Polynomial and Function: Parabola
as shown in the following Figure 4.19:

Figure 4.19 OriginPro Parabola Curve Fitting

We could find the parameter values of A, B, C, R2 and F values as

shown in the following Figure 4.20:

Figure 4.20 OriginPro Parabola Curve Fitting R2Values

 89

By scrolling the window it could be found the parabola fitted

curves plot as shown in the following Figure 4.21:

Figure 4.21 OriginPro Final Parabola Curve Fitting Diagram

By double-clicking the curve window, we could get the enlarged

curve as shown in the following Figure 4.22for our analysis purpose to find
best fit curve.

Figure 4.22 OriginPro Enlarged Final Parabola Curve Fitting Diagram

 90

OriginPro software has created the data analyses for POME ink
based on quadratic equation under parabola the curve fitting was performed
by the nonlinear least squares fitter which is based on the Levenberg–
Marquardt algoriwahm. The curve fitting data would be further used for
mathematical analysis to verify the consistent results of POME ink
performance on four color offset machine. The results in terms of quality on
gloss coated, matte coated and uncoated papers have been were scientifically
assessed.

4.3 MATHEMATICAL MODELING

Mathematical modeling describes the different aspects of the real

world, their interaction, and their dynamics through mathematics.
Mathematical models also offer new possibilities to manage the increasing
complexity of technology. In this research, a mathematical model is derived
to have a deeper understanding of the relationship between the ink density and
color values which determine the print quality. The ink density 6D for
different values of 6Eab were calculated for determining a mathematical
relationship between them to predict the value of 6Eab based on the value of
6D, and explain the impact of changes in the values 6Eab on the values of
6D.

Curve fitting is the process of constructing a curve, or mathematical

functionthat has the best fit to a series of datapoints possibly subject to
constraints. Curve fitting can involve either interpolation, where an exact fit
to the data is required, or smoothing, in which a "smooth" function is
constructed that approximately fits the data. A related topic is regression
analysis, which focuses more on questions of statistical inference such as how
much uncertainty is present in a curve that is fit to data observed with random
errors. Fitted curves can be used as an aid for data visualization, to infer
values of a function where no data are available, and to summarize the
 91

relationships among two or more variables. Extrapolation refers to the use of

a fitted curve beyond the range of the observed data, and is subject to a degree
of uncertaintysince it may reflect the method used to construct the curve as
much as it reflects the observed data.

4.3.1 Regression Model

A regression model is applied to find the relationship between the

ink density and color values. Out of the two variables density variation 6D
and color difference 6Eab, 6Eab is considered as the dependent variable and
6D as the independent variable. Using regression model, it is studied that
changes in the values of dependent variable are resulted by the changes in the
values of independent variable. Regression model polynomial parabola curve
fit is used in this research to arrive the mathematical model was shown in
Figure 4.23.

Figure 4.23 Parabola Curve fit

The relationship between 6D and 6Eab is represented by the

polynomial parabola fit is given by 6Eab = a + b6D + c6D2. Fitting a model
 92

means obtaining the values of the parameters a, b and c on the basis of

collected observations on 6D and 6Eab.

4.3.2 Modeling Tool (originPro)

The OriginPro 8 SR4 version software is used in this research for

curve fitting using regression analysis to find the relationship between ink
density 6D and color parameters 6Eab. Origin is an industry-leading
scientific graphing and data analysis GUI software developed by OriginLab®.
Curve fitting is one of Origin’s most powerful and most widely used
analytical methods. Origin provides tools for linear and nonlinear curve
fitting. The output data of printed sheet results are assessed by feeding the
readings of Cyan, Magenta, Yellow and Black density readings 6D and 6Eab
color differences to find the best fit regression model.

4.4 GOODNESS OF FIT

A regression model is first developed, and then the best fit

parameters are estimated using the least-square method. Finally, the quality of
the model is assessed using hypothesis tests of R2, T-value and F-value of
ANOVA to ascertain the fitness of the regression model. In X axis the 6D
optical density and Y axis 6Eab color parameters are taken, the readings were
fitted in the regression line.

Linear regression calculates an equation that minimizes the distance

between the fitted line and all of the data points. Technically, ordinary least
square regression minimizes the sum of the squared residuals. In general, a
model fits the data well if the difference between the observed values and the
model’s predicted values are small and unbiased. The fitness of the model can
be proved using the following parameters:
 93

1. Coefficient of determination (R2)

2. t-test statistics

3. F-test (ANOVA)

4.4.1 Coefficient of Determination

The goodness of the fitted model is studied by calculating the

Coefficient of determination R2. Statistical measure R2 is the closeness of the
data to the fitted regression line. R2 is the percentage of the response variable
variation that is explained by a linear model.

R2 = Explained variation / Total variation

)n (yi — ȳ)2
R2 = )ni=1(yi — ȳs )2
i=1

R2 is always between zero and one. Zero indicates that the model
explains none of the variability of the response data around its mean and one
indicates that the model explains all the variability of the response data
around its mean. In general, the higher the R-square, the better the model fits
the data.

4.4.2 t-Test for the Slope

The t-value is a statistic test that measures the difference between

an observed sample statistic and its hypothesized population parameter in
units of standard error. A t-test compares the observed t-value to a critical
value on the t-distribution with (n-1) degrees of freedom to determine whether
the difference between the estimated and hypothesized values of the
population parameter is statistically significant.
 94

For the fitted model, the hypothesis is tested. The slope is zero
against H0: b=0 against Ha: bÇ0 is tested. The test statistics is

b̂
Tc =
)n
i=1(yi – ŷ)
2
J (n–2) )n (si – s)¯ 2
i=1

The T-statistics value at 5 % level of significance and 13 degree of

freedom, the critical value referred from the t- distribution Table minimum is
2.160. In general, the higher value of 2.160, the better the model fits the data.

4.4.3 F-Test (ANOVA)

F-test can be used as an alternative for t-test for linearity. In F-test

also, the hypothesis H0: b = 0 is tested. This test is based on the concept of
partitioning the total variability into explained variability and unexplained
variability.

)n (ŷs — ȳ)2 x
F = )i=1
n
(yi — ŷ)2
i=1

An F-test is any statistical test in which the test statistic has an F-

distribution under the null hypothesis. It is most often used when comparing
statistical models that have been fitted to a data set, in order to identify the
model that best fits the population from which the data were sampled. Exact
“F-tests” mainly arise when the models have been fitted to the data using least
squares. F-value at 5% level of significance and (1, 13) degree of freedom, the
critical value using F distribution Table is 4.67. In general, the higher value of
4.67, the better the model fits the data.
 95

4.5 CLOSURE

The POME ink print results have been analyzed with mathematical
regression model and found that there is a linear relationship between printing
ink density 6D and color density variation 6Eab of various substrates. The
experimental results demonstrate that the LAB values of POME ink is
desirable on various substrates and it has matched with ISO 12647-2.