Summary  
This chapter explains how a Scikit-learn estimator implements the core methods—`.fit()` for training, `.predict()` for generating predictions, and `.score()` for evaluating performance.  

General domain of usage  
Supervised machine learning (e.g., classification)

The fundamentals of data preprocessing and pipeline construction are now covered. The next step is **modeling**.


A **model** in Scikit-learn is an **estimator** that provides `.predict()` and `.score()` methods, along with `.fit()` inherited from all estimators.


## .fit() 

Once the data is preprocessed and ready to go to the model, the first step of building a model is **training a model**. This is done using the `.fit(X, y)`.

For **supervised learning** (regression, classification), `.fit()` requires both `X` and `y`.
For **unsupervised learning** (e.g., clustering), you call `.fit(X)` only. Passing `y` does not cause an error — it is simply ignored.

Note

During training, the model **learns** patterns needed for prediction. What it learns and how long training takes depend on the algorithm. Training is often the **slowest part** of ML, especially with large datasets.

## .predict()

After training, use `.predict()` to generate predictions:

```python
model.fit(X, y)
y_pred = model.predict(X_new)
```

## .score()

`.score()` evaluates a trained model, typically on a **test set**:

```python
model.fit(X, y)
model.score(X_test, y_test)
```

It compares predictions with true targets. By default, the metric is **accuracy** for classification.

`X_test` refers to the subset of the dataset, known as the **test set**, used to evaluate a model's performance after training. It contains the **features** (input data). `y_test` is the corresponding subset of **true labels** for `X_test`. Together, they assess how well the model predicts new, unseen data.

Machine learning drives modern technological innovation across all industries. Embark on a comprehensive introduction to predictive modeling by mastering foundational algorithmic concepts utilizing Scikit-Learn. Participants will construct robust classifiers, evaluate performance metrics, ultimately culminating in the development of a complete predictive project.

Models

.fit()

.predict()

.score()