Solving problems with stochastic optimization algorithms: hill climbing, simulated annealing, genetic algorithm.
Sample problems: best travel schedule, optimal student-to-dorm assignments.
Extracting linear and polynomial models from data using regression and logistic regression to predict and classify.
Sample problems: predicting housing prices, classifying images as cats/dogs.
Decision trees break data into groups along different attribute values and classify each group separately. Discovers important features on top of the tree. Variation: Decision Table based on classification rules. Regression trees predict numeric targets.
Sample problems: predicting Titanic survival, predicting COVID outcome based on preconditions.
Instead of building predictive models from data, these classifiers just remember all the data and classify new instance using the Nearest Neighbor approach: find most similar instances and issue their class as a prediction. Also used for recommenders: find similar users and recommend their favorite items.
Sample problems: housing prices revisited.
Algorithms for grouping observation into clusters based on pairwise similarity between records. Unsupervised learning: discover clusters which are unknown classes of objects.
Sample problems: document clustering, finding best score locations on the map.
This unsupervised technique is used to discover groups of items which often appear together. Also used to make recommendations.
Sample problem: market basket analysis.