Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Regression Modelling

Simple Linear Models


1. Introduction
Lecture 1 2. Assumptions
3. Parameter Estimation

Karanjah Anthony Ph.D.


Multimedia University
FoST-Department of Mathematics
2022@tkaranjah 1 2022@tkaranjah 2

Overview Overview
Many problems in sciences involve a However, in many situations, the relationship
study or analysis of the relationship between variables is not deterministic.
between two or more variables. For example, the electrical energy consumption
For example; of a house (y) is related to the size of the house
◼ Income and number of years of education (x), in square feet), but it is unlikely to be a
deterministic relationship.
◼ Height and weight of people
◼ the pressure of a gas in a container is related Similarly, the fuel usage of an automobile (y) is
to the temperature, related to the vehicle weight (x), but the
◼ Dose of a drug and response relationship is not a deterministic one.

2022@tkaranjah 3 2022@tkaranjah 4

Overview Overview
In both of these examples, the value of the The collection of statistical tools that are used
response of interest y (energy consumption, to model and explore relationships between
fuel usage) cannot be predicted perfectly from variables that are related in a non-deterministic
knowledge of the corresponding x. manner is called regression analysis.

It is possible for different automobiles to have Because problems of this type occur so
different fuel usage even if they weigh the frequently in many branches of science and
same, and it is possible for different houses to engineering, regression analysis is one of the
use different amounts of electricity even if they most widely used statistical tools
are the same size.

2022@tkaranjah 5 2022@tkaranjah 6

1
Overview Simple Linear Regression
For example, in a chemical process, suppose that The case of simple linear regression considers a
the yield of the product is related to the process- single regressor variable or predictor variable x
operating temperature. and a dependent or response variable Y.

Regression analysis can be used to build a model


to predict yield at a given temperature level. Suppose that the true relationship between Y and
x is a straight line and that the observation Y at
This model can also be used for process each level of x is a random variable. As noted
optimization, such as finding the level of previously, the expected value of Y for each
temperature that maximizes yield, or for process value of x is
control purposes.

2022@tkaranjah 7 2022@tkaranjah 8

Simple Linear Regression Simple Linear Regression


◼ ◼

2022@tkaranjah 9 2022@tkaranjah 10

Simple Linear Regression Simple Linear Regression


◼ ◼

2022@tkaranjah 11 2022@tkaranjah 12

2
Simple Linear Regression
Assumptions Simple Linear Regression

2022@tkaranjah 13 2022@tkaranjah 14

Simple Linear Regression



Simple Linear Regression

2022@tkaranjah 15 2022@tkaranjah 16

Simple Linear Regression Simple Linear Regression


an and the sum of the squares of the deviations of
the observations from the true regression line is

2022@tkaranjah 17 2022@tkaranjah 18

3
Simple Linear Regression Simple Linear Regression
The least squares estimates of the intercept and
slope in the simple linear regression model are;

2022@tkaranjah 19 2022@tkaranjah 20

Simple Linear Regression Simple Linear Regression


Later, we will use the residuals to provide
information about the adequacy of the fitted
model.

Deviations of the data from the estimated


regression model are can be represented
diagrammatically as follows

2022@tkaranjah 21 2022@tkaranjah 22

Simple Linear Regression Simple Linear Regression


Notationally, it is occasionally convenient to give Example
special symbols to the numerator and For example, in a chemical process, suppose that
denominator of above Equation. the yield of the product y is related to the
process-operating temperature x.
Specifically the data in the table below table, y is
the purity of oxygen produced in a chemical
distillation process, and x is the percentage of
hydrocarbons present in the main condenser of
the distillation unit.

Required: Find the parameter estimates.

2022@tkaranjah 23 2022@tkaranjah 24

4
Simple Linear Regression Simple Linear Regression

2022@tkaranjah 25 2022@tkaranjah 26

Simple Linear Regression Simple Linear Regression


We will fit a simple linear regression model to the
oxygen purity data in Table above.
The following quantities may be computed;

2022@tkaranjah 27 2022@tkaranjah 28

Simple Linear Regression Simple Linear Regression


Therefore, the least squares estimates of the This model is plotted in Fig. below, along with the
slope and intercept are; sample data

The fitted simple linear regression model (with the


coefficients reported to three decimal places) is;

2022@tkaranjah 29 2022@tkaranjah 30

5
Simple Linear Regression Simple Linear Regression
Practical Interpretation:

The value of which 14.9475 is interpreted as follows:

When when the hydrocarbon level changes by1.00% then purity


increases by 14.9475 units.

2022@tkaranjah 31 2022@tkaranjah 32

Simple Linear Regression


Practical Interpretation:

What is the interpretation of ?

END

2022@tkaranjah 33 2022@tkaranjah 34

You might also like