Day-13 – Hyperparameter Tuning + GridSearchCV

 Today you learn:

How to improve ML model performance systematically 🚀

This is VERY important for:

• Interviews
• Real projects
• Kaggle competitions
• Production ML

🎯 Goal of Day-13

You will:

✅ Understand hyperparameters
✅ Tune ML models
✅ Use GridSearchCV
✅ Find best model settings

🧠 What is Hyperparameter?

Simple meaning:

Settings you choose BEFORE training model

Example: 

RandomForestClassifier(n_estimators=100)

👉 n_estimators is hyperparameter.

Model learns:

weights, patterns

But YOU choose:

number of trees
depth
learning rate

🧠 Why Hyperparameter Tuning Matters

Same model:

Setting	Accuracy
Default	75%
Tuned	90%

👉 Better settings = better model.

🚀 Part 1 – Import Libraries

import pandas as pd

from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score

🚀 Part 2 – Create Dataset

data = {
    "Hours": [1,2,3,4,5,6,7,8],
    "Marks": [40,45,50,55,70,80,90,95],
    "Pass":  [0,0,0,0,1,1,1,1]
}

df = pd.DataFrame(data)

X = df[["Hours", "Marks"]]
y = df["Pass"]

🚀 Part 3 – Train/Test Split

X_train, X_test, y_train, y_test = train_test_split(
X,
y,
test_size=0.2,
random_state=42

)

🚀 Part 4 – Create Model

model = RandomForestClassifier()    

🚀 Part 5 – Define Hyperparameters

params = {
"n_estimators": [5, 10, 20],
"max_depth": [2, 4, 6]
}

🧠 Meaning

Parameter	Meaning
n_estimators	Number of trees
max_depth	Tree depth

🚀 Part 6 – GridSearchCV

grid = GridSearchCV(
estimator=model,
param_grid=params,
cv=3
)

🧠 What Happens?

GridSearch tries ALL combinations:

5 trees + depth 2
5 trees + depth 4
10 trees + depth 2
...

Then selects BEST model.

🚀 Part 7 – Train Grid Search

grid.fit(X_train, y_train)

🚀 Part 8 – Best Parameters

print("Best Params:", grid.best_params_)

🚀 Part 9 – Best Model Prediction

best_model = grid.best_estimator_

y_pred = best_model.predict(X_test)

accuracy = accuracy_score(y_test, y_pred)

print("Accuracy:", accuracy)

🧠 Real AI Insight

Without tuning:

average model

With tuning:

optimized model

🧠 Important Concept – Cross Validation

In GridSearch:

cv=3

Means:

• Train/test multiple times
• More reliable evaluation

⚠ Important Interview Question

Q:

What is GridSearchCV?

Answer:

A technique to automatically test multiple hyperparameter combinations and select the best model using cross-validation.

🧠 Real AI Insight

This is used heavily in:

• Kaggle competitions
• Production ML
• Enterprise AI optimization

🎯 End of Day-13 Goals

You now:

✅ Understand hyperparameters
✅ Tune ML models
✅ Use GridSearchCV
✅ Improve performance systematically