Supervised Machine Learning in Python

A practical course about supervised machine learning using Python programming language

In this practical course, we are going to focus on supervised machine learning and how to apply it in Python programming language.

Supervised machine learning is a branch of artificial intelligence whose goal is to create predictive models starting from a dataset. With the proper optimization of the models, it is possible to create mathematical representations of our data in order to extract the information that is hidden inside our database and use it for making inferences and predictions.

A very powerful use of supervised machine learning is the calculation of feature importance, which makes us better understand the information behind data and allows us to reduce the dimensionality of our problem considering only the relevant information, discarding all the useless variables. A common approach for calculating feature importance is the SHAP technique.

Finally, the proper optimization of a model is possible using some hyperparameter tuning techniques that make use of cross-validation.

With this course, you are going to learn:

1. What supervised machine learning is

2. What overfitting and underfitting are and how to avoid them

3. The difference between regression and classification models

4. Linear models

   1. Linear regression

   2. Lasso regression

   3. Ridge regression

   4. Elastic Net regression

   5. Logistic regression

5. Decision trees

6. Naive Bayes

7. K-nearest neighbors

8. Support Vector Machines

   1. Linear SVM

   2. Non-linear SVM

9. Feedforward neural networks

10. Ensemble models

    1. Bias-variance tradeoff

    2. Bagging and Random Forest

    3. Boosting and Gradient Boosting

    4. Voting

    5. Stacking

11. Performance metrics

    1. Regression

       1. Root Mean Squared Error

       2. Mean Absolute Error

       3. Mean Absolute Percentage Error

    2. Classification

       1. Confusion matrix

       2. Accuracy and balanced accuracy

       3. Precision

       4. Recall

       5. ROC Curve and the area under it

       6. Multi-class metrics

12. Feature importance

    1. How to calculate feature importance according to a model

    2. SHAP technique for calculating feature importance according to every model

    3. Recursive Feature Elimination for dimensionality reduction

13. Hyperparameter tuning

    1. k-fold cross-validation

    2. Grid search

    3. Random search

All the lessons of this course start with a brief introduction and end with a practical example in Python programming language and its powerful scikit-learn library. The environment that will be used is Jupyter, which is a standard in the data science industry. All the Jupyter notebooks are downloadable.