TUe (Eindhoven University of Technology) Assignments

Code written for some assignments made as part of my Master studies at TU Eindhoven between September 2018 and July 2020.

Multi-Layer Perceptron (MLP)

The notebook "Multi-Layer Perceptron" contains the code developed for creating a MLP Neural Network manually without the use of libraries. It is a good exercise for understanding how a MLP works step by step and improve one's intuition on what this algorithm does.

The code was written by myself as partial fulfillment for the course Foundations of Data Mining at TU Eindhoven.

Deep Learning & Recommender Systems

Contains code written for assigments for the course Deep Learning and Recommender Systems in TU Einhdhoven. Most of the work is related to Convolutional Neural Networks.

CNN for Image Classification

In this notebooks, some CNN are trained with the purpose of Image classification.

AutoEncoders

In this notebook we first use the last layer of a CNN to extract image features and then use KNN for determining the most similar images to any given image. Later, AutoEncoders are trained with the purpose of denoising noisy images.

Siamese Networks

In this notebook, Siamese Networks are trained for the purpose of determining whether any two given images belong to the same class or not.