Module 7 Lesson 3: Scikit-Learn: The ML Toolkit

In the world of Python AI, Scikit-Learn (Sklearn) is the undisputed champion. It contains almost all the classic Machine Learning algorithms you'll ever need. The best part? Every algorithm in Sklearn works the exact same way. Once you learn the pattern for one, you've learned them all.

Lesson Overview

In this lesson, we will cover:

• The Sklearn Pattern: Import, Instantiate, Fit, Predict.
• Handling Data: X (Features) and y (Target).
• The Train/Test Split: Why you must never test on data the model has already seen.
• Your First "Hello AI": A conceptual code walkthrough.

---

1. The 4-Step Pattern

Building an AI model in Scikit-Learn always follows these four steps:

1. Import: Bring in the algorithm from the library.
2. Instantiate: Create an "object" of that model (like we did in Module 4!).
3. Fit: Train the model on your data (The "Learning" phase).
4. Predict: Use the model to guess answers for new data.

---

2. Setting Up the Data

In the industry, we use X (Uppercase) for Features and y (Lowercase) for the Target.

# Square Footage (X), Price (y)
X = [[1000], [1500], [2000]] 
y = [200000, 300000, 400000]

---

3. The Train/Test Split

Imagine a student who memorizes the answers to an exam. If you give them the exact same questions, they'll get 100%, but they haven't actually learned anything. If you give them new questions, they might fail.

To prevent this in AI, we split our data:

• Training Set (80%): Used for the model to learn.
• Testing Set (20%): Used to see if the model actually understood the patterns.

from sklearn.model_selection import train_test_split

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

---

4. Why Consistency is Key?

Because Sklearn is so consistent, if you decide that a "Linear Regression" model isn't working well, you can swap it for a "Decision Tree" model by changing just one line of code. The fit() and predict() methods stay exactly the same.

---

Practice Exercise: The Pattern Reciter

Without writing the full code yet, write down the 4 steps of the Scikit-Learn pattern and describe what happens in the "Fit" phase. Why do we call it "Training"?

---

Quick Knowledge Check

1. What is the name of the most popular Python library for Machine Learning?
2. What does the fit() method do?
3. Why do we split data into "Train" and "Test" sets?
4. What is the conventional variable name for the target (the thing we want to predict)?

---

Key Takeaways

• Scikit-Learn provides a unified interface for hundreds of ML algorithms.
• The "Import -> Instantiate -> Fit -> Predict" workflow is universal.
• Proper data splitting is the only way to measure a model's true accuracy.
• Python's OOP structure (Module 4) is what makes this consistent interface possible.

---

What’s Next?

It’s time to build your first working AI. In Lesson 4, we’ll use the 4-step pattern to create a Linear Regression model that predicts house prices!