Читать книгу Data Analytics in Bioinformatics - Группа авторов - Страница 23

Linear and Logistics regression are two common types of regression used for prediction. The result of the prediction is represented with the help of numeric variables. The difference between linear and logistic regression is depicted in Table 1.3 for easy understanding.

Linear regression is used to model the data by using a straight line whereas the logistic regression deals with the modeling of probability of events in a bi-variate manner that is occurring as a linear function of dependent variables. Few other types of regression analysis are depicted by different scientists and listed below.

Table 1.3 Difference between linear & logistic regression.

S. No.	Parameter	Linear regression	Logistic regression
1	Purpose	Used for solving regression problems.	Used for solving classification problems.
2	Variables Involved	Continuous Variables	Categorical Variables
3	Objective	Finding of best-fit-line and predicting the output.	Finding of s-curve and classifying the samples.
4	Output	Continuous Variables such as age, price, etc.	Categorical Values such as 0 & 1, Yes & No.
5	Collinearity	There may be collinearity between independent attributes.	There should not be collinearity between independent attributes.
6	Relationship	The relationship between a dependent variable and the independent variable must be linear.	The relationship between a dependent variable and an independent variable may not be linear.
7	Estimation Method Used	Least Square Estimation	Maximum Likelihood Estimation

 Polynomial Regression: It is used for curvilinear data [57–58].

 Stepwise Regression: It works with predictive models [59–60].

 Ridge Regression: Used for multiple regression data [61–62].

 Lasso Regression: Used for the purpose of variable selection & regularization [63–64].

 Elastic Net Regression: Used when the penalties of lasso and ridge method are combined [65].