Machine learning in Python

Exercises and assignments from a machine learning course in my fourth semester. It goes from linear regression all the way to neural networks, unsupervised learning and genetic algorithms, mostly using scikit-learn, NumPy, pandas and matplotlib.

Contents

Folder	Topic
Lektion01	Python warm-up exercises
Lektion02	Linear regression, including polynomial features and 3D plots
Lektion03	Genetic algorithms: cannonball trajectory, flight scheduling, healthcare scheduling
Lektion04	Logistic regression and classification metrics (precision/recall on the Iris dataset)
Lektion05	Data exploration and simulation: Chicago crime data, world population, SIR infection model, Snake
Lektion06	Decision trees, random forests, SVMs and k-means (incl. the moons dataset)
Lektion07	SVM on breast-cancer data, plus a small pygame tic-tac-toe game
Lektion08	Multi-layer perceptrons, diabetes dataset
Lektion09	MLP on the moons dataset, Titanic survival prediction
Lektion10	MNIST digits, k-means and PCA on Iris
Lektion11	Travelling salesman problem with a genetic algorithm (1000 cities)
Lektion12	Predicting Pokémon type from stats with a random forest
Mandatory_Assignment	Graded assignments: neural nets on moons/circles, Titanic survival, face recognition with PCA + SVM

Files named *_Min_Løsning.py ("my solution") are my own answers to the assignment snippets.

Tech

Python 3 · scikit-learn · NumPy · pandas · matplotlib · pygame

Running it

pip install scikit-learn numpy pandas matplotlib pygame
cd Lektion06
python decisiontree.py

Scripts read their datasets (CSV/TXT) from their own folder, so run them from inside that folder.