This repository contains the homeworks for the course 2AMM15 (Machine Learning Engineering) at Eindhoven University of Technology. The assignments cover various topics and techniques in machine learning engineering.
- Homework 1 - Develop a machine learning model for recognizing Devanagari script characters. Enhance the model's performance using image embeddings and perform model inspection, including confusion matrix analysis, to identify and understand misclassification patterns.
- Homework 2 - Build predictive models for real-world data from ESA's Mars Express spacecraft. Focus on forecasting energy influx and power line currents, and incorporate domain-specific knowledge using descriptions generated by large language models (LLMs).
- Homework 3 - Design and implement a convolutional neural network (CNN) and a transfer learning model using MobileNet for indoor scene recognition. Collect and use images to create a new dataset for testing the model and evaluate its performance.