SUPERVISED LINEAR REGRESSION USING OLS MODEL

REGRESSION ANALYSIS

Purpose of Regression analysis

LINEAR REGRESSION

Variance
COVARIANCE -TOGETHER SPREAD OF X AND Y
CORRELATION (R): COV(x)/SIGMA (x) IS ZSCALED FORMULA SO IT IS DOING SCALING

Linear Regression two methods as shown below:

ORDINARY LEAST SQUARE

Simple Linear Regression — one independent and one dependent variable.

Random error component:

OLS OBJECTIVE

BEST FIT LINE

INTERPRETATION OF BETA COEFFICIENTS (b0 and b1):

slope (b1)
Intercept (b0)

MEASURES OF VARIATION:

#OLS MODEL CODE

Explainability

MATH BEHIND OLS

LOSS FUNCTION / COST FUNCTION/ERROR FUNCTION

MODEL EVALUATION METRICS

Rsquared and Adjusted Rsquared:

F-STATISTIC

OPTIMIZATION-GRADIENT DESCENT

LOSS FUNCTION / COST FUNCTION/ERROR FUNCTION

MEAN SQUARED FUNCTION

GRADIENT DESCENT ALGORITHM

Data scientist Aspirant passionate in learning new technologies and sharing my thoughts to others .