Machine Learning Course - TDDE01

This repository contains lab-solutions for the TDDE01 Machine Learning course taken at Linköping University during the fall of 2023. The course includes three labs focusing on core ML concepts.

Contents

1. Lab Descriptions
2. How to Run
3. Resources

Lab Descriptions

Lab 1: Introduction to ML

• KNN: Classify handwritten digits (kknn package).

• Linear Regression: Predict Parkinson’s metrics.
• Logistic Regression: Binary classification with basis expansion.

Lab 2: Advanced Models

• Regularization: LASSO and Ridge regression on Tecator data.

• Decision Trees: Tree-based prediction and analysis.

• PCA: Dimensionality reduction on crime data.

Lab 3: Kernel Methods, SVMs, Neural Networks

• Kernel Methods: Predict temperatures using Gaussian kernels.

• SVM: Spam classification (kernlab package).

• Neural Networks: Learn sine function with various activations.

How to Run

• Set Up: Install required R packages: kknn, caret, glmnet, kernlab, geosphere, neuralnet.
• Execution: Run .r scripts in respective lab folders.
• Reproducibility: Use set.seed(12345) for consistent results.

Resources

• Helpfile.pdf: Located in /Other/, containing exam tips.
• data-wrangling-cheatsheet.pdf: Data-wrangling with dplyr and tidyr made easy.