Statistics
Class 11-12

Linear Regression

y^=β0+β1x\hat{y} = \beta_0 + \beta_1 x

What is this? (Explained Simply)

Imagine your class took a math test. You want to know: do students who study MORE hours get BETTER marks? You plot each student as a dot (hours studied vs marks). The dots are scattered, but you can see a rough pattern going upward. Linear regression draws the BEST straight line through those scattered dots. It's like drawing a line that gets as close as possible to all the dots. Now you can predict: 'If I study 5 hours, I'll probably get around 80 marks!'

024681005101520
Data PointsBest Fit Line

Adjust Variables

Slope
slope =
0.15
Intercept
intercept =
-55
Noise Level
noise =
05

Linear regression finds the best straight line through scattered data points. It minimizes the total distance between the line and all data points.

Real-World Applications

Predicting exam marks: If students who study 2 hrs score 60 and those studying 4 hrs score 80, regression predicts marks for any study hours

Real estate: Predicting house prices based on area in sq. feet

Sales forecasting: Predicting next month sales based on advertising spend

What would an intelligent skeptic say?

Regression finds correlation, not causation. Ice cream sales and drowning deaths are strongly correlated (both rise in summer). A regression line would suggest buying ice cream causes drowning. Also, regression always finds a 'best fit' line — even for random noise. R-squared of 0.3 means 70% of the variation is UNEXPLAINED. Most published regression results are weaker than they appear.

Community Explanations

No community explanations yet. Be the first to share yours!

to write your own explanation

Community Aha! Moments

to share your insights

No insights yet. Be the first to share!

Related Formulas

Great exploration of Statistics!

With Pro, you get AI explanations that connect formulas to real-world applications.

Ask AI to explain any formula
Real-world connections

Install i See f(x)

Add to home screen for quick access