Linear Regression
Presented By
AMITHYA SANTHOSH
preencoded.png
Understanding the Linear Regression Model
What is it? Key Components
Linear Regression predicts a dependent variable's value based Independent variables (predictors) influence the dependent
on the relationship with one or more independent variables. It variable (outcome). The regression equation describes the
assumes a linear relationship between these variables. The relationship between them. The slope and intercept
model aims to find the line of best fit that minimizes the determine the line's shape.
distance between data points and the line.
preencoded.png
Assumptions of Linear
Regression
Linearity Independence
The relationship between Data points should be
independent and dependent independent of each other.
variables should be linear.
Homoscedasticity Normality
The variance of errors should Errors should be normally
be constant across all values distributed.
of independent variables.
preencoded.png
Simple Linear Regression: Fitting the Model
Data Preparation
1 Gather relevant data and ensure it meets the assumptions.
Model Estimation
2 Use statistical software to estimate the coefficients and intercept of the
regression equation.
Model Interpretation
3 Analyze the coefficients to understand the relationship
between the variables.
preencoded.png
EQUATIONS
y = mx + c
Slope: 𝑚 = 𝑁 ( ∑ 𝑥 𝑦 ) − ( ∑ 𝑥 ) ( ∑ 𝑦 )
𝑁 ( ∑ 𝑥 ² ) − ( ∑ 𝑥 )²
Intercept : 𝑐 = ( ∑ 𝑦 ) − 𝑚 ( ∑ 𝑥 )
𝑁
preencoded.png
Multiple Linear Regression:
Handling Multiple Variables
Multiple Predictors Complex Relationships
Extends the model to incorporate Enables analyzing the effects of
multiple independent variables. different variables on the outcome.
Model Complexity
Requires careful consideration of
variable interactions and model
selection.
preencoded.png
Applications of Linear Regression: Predictive
Modeling
Business
1
Predicting sales, customer churn, and market trends.
Finance
2
Forecasting stock prices, analyzing investment performance.
Healthcare
3
Estimating patient risk, predicting disease progression.
preencoded.png
Limitations and Considerations
in Linear Regression
1 2
Non-linearity Outliers
Linear Regression may not accurately Outliers can heavily influence model
capture non-linear relationships. results.
3
Multicollinearity
High correlation between independent
variables can lead to unstable results.
preencoded.png