What is Linear Regression?
Imagine you’re running a small chai shop, and you start noticing something:
The more hours your shop stays open, the more cups of chai you sell.
Now you start wondering:
“Can I predict how many cups I’ll sell if I keep the shop open for, say, 10 hours?”
This is where Linear Regression comes in.
Linear Regression is a method to:
Find the relationship between two things– one that you control (input) and one that you observe (output) and then use that relationship to make predictions.
Real-Life Example: Chai Shop
Let’s say you collect this data over a week:
| Hours Open | Cups Sold |
|---|---|
| 2 hours | 20 cups |
| 4 hours | 40 cups |
| 6 hours | 60 cups |
| 8 hours | 80 cups |
You can clearly see a pattern:
Each hour = 10 extra cups sold.
If we draw this on a graph, the points fall in a straight line. That line is called the regression line.
Prediction:
Now, if someone asks you:
“If I stay open for 10 hours, how many cups will I sell?”
You say:
Cups Sold = 10 × Hours Open
Cups Sold = 10 × 10 = 100
That’s Linear Regression using past data to draw a line and predict the future.
In Technical Terms
- Input (X): Hours Open
- Output (Y): Cups Sold
- Linear Equation:
Y = mX + cwhere:mis the slope (how much Y changes when X increases)cis the intercept (value of Y when X = 0)
In our case:
m = 10c = 0(If shop isn’t open, no chai is sold)
Where Is Linear Regression Used?
- Predicting house prices (based on size)
- Estimating salary (based on experience)
- Forecasting sales (based on ad spend)
- Health: Predicting blood pressure from age
🧑💻 Code Example
from sklearn.linear_model import LinearRegression
# Data
hours = [[2], [4], [6], [8]]
cups = [20, 40, 60, 80]
# Model
model = LinearRegression()
model.fit(hours, cups)
# Predict
model.predict([[10]]) # Output: [100.]
Summary
- Linear Regression finds a line that best fits your data.
- It’s great when your data shows a linear trend.
- It helps in making predictions using simple math.
- Think of it like drawing the best-fit line through a scatter plot.