Master data preprocessing and modeling techniques with scikit-learn. Build machine learning models like linear regression and decision trees, focusing on data training, testing, and evaluation.

Become a Machine Learning Engineer

In this module, you'll explore preprocessing data using scikit-learn, including techniques such as feature extraction, scaling, discretizing, encoding, dimensionality reduction, and handling missing values. You'll then explore building and implementing machine learning models like linear regression, logistic regression, and decision trees. The module also covers data training, data testing, cross-validation, and model evaluation methods to ensure the accuracy and robustness of your machine learning solutions.

Preprocessing Data

When we use `MinMaxScaler` to scale a feature, what will be its range?

When we use <code>MinMaxScaler</code> to scale a feature, what will be its range?

When we use `StandardScaler` to scale a feature, what will be its range?

When we use <code>StandardScaler</code> to scale a feature, what will be its range?

Mean = $0$ \
Standard deviation = same as the original feature

Mean = same as the original feature \
Standard deviation = $1$

When we use `StandardScaler` to scale a feature, what will be its mean and standard deviation?

When we use <code>StandardScaler</code> to scale a feature, what will be its mean and standard deviation?

If we want to discretize a continuous variable in a way that each class has the same number of observations, which class should we use?

If we want to discretize a continuous variable in a way that each class has the same number of observations, which class should we use?

Imagine we have a categorical column that we want to turn numerical, while keeping it in one column. Which encoder should we use?

Imagine we have a categorical column that we want to turn numerical, while keeping it in one column. Which encoder should we use?

If our original dataset has 10 features and we apply PCA to it in the following way:
```python
# Perform PCA
pca = PCA(n_components=2)
X_pca = pca.fit_transform(X)
```
How many columns would we get at the end?

If our original dataset has 10 features and we apply PCA to it in the following way:
<pre><code class="language-python"># Perform PCA
pca = PCA(n_components=2)
X_pca = pca.fit_transform(X)
</code></pre>
How many columns would we get at the end?

Quiz: Preprocessing

Quiz: Preprocessing

Check your understanding of preprocessing and its steps.

Set up for Machine Learning with scikit-learn

Preprocessing with Scikit-Learn

Modeling with Scikit-Learn

Conclusion

Quiz: Preprocessing