This is where the math and data skills from the earlier modules turn into working models. Machine learning is about finding patterns in data and using them to make predictions. This module covers the classic algorithms that still power most real-world systems: linear and logistic regression, decision trees, random forests and boosting, support vector machines, k-nearest neighbors, and clustering.
You'll also learn how to judge a model: splitting data, spotting overfitting and underfitting, balancing bias and variance, and choosing the right metrics. The later lessons move into scikit-learn, the standard Python library for classical machine learning, and finish with end-to-end projects, from a first model to deploying one.
These courses are long, so treat them as a reference: watch the parts you need, then come back for the rest. When you're done, continue on to Deep Learning.
The basics for absolute beginners: features, classification vs regression, k-nearest neighbors, Naive Bayes, logistic and linear regression, support vector machines, a first neural network in TensorFlow, then k-means clustering and PCA.
The ML workflow, linear and logistic regression with regularization, SVMs, PCA, learning theory and the bias-variance tradeoff, decision trees, ensembles and boosting, clustering, plus projects like house price prediction and spam detection.
An overview of scikit-learn, preprocessing tools, metrics and custom metrics for judging models, meta estimators, and the human-learn library.
Hands-on and project-driven: linear regression and gradient descent, logistic regression, decision trees and random forests, XGBoost, unsupervised learning, a project built from scratch, and deploying a model with Flask.