kennardmah / MLforDesEng

Compilation of coursework and notes for the "Machine Learning for Design Engineers" course at Imperial

MLforDesEng

This repository contains notes for the "Machine Learning for Design Engineers" module offered at Imperial College London. The module is assessed in three categories: two written reports (15%), lab exam (15%), and final exam (70%). In this repo, you'll find my notes for the course as well as an in-depth analysis of each labs to prepare for the lab exam. You can also find the written reports attached.

Lab 1 - Building an ANN (Artificial Neural Network) with a sigmoid activation function and backpropagation

1. Understand and code a simple neuron
2. Understand how a neuron learns
3. Understand its limitations

Lab 2 - Adding Non-linearity (Hidden layers) to solve XOR problems with softmax activation and cross-entropy loss function

1. Understand Backpropagation (Deriving each partial derivatives)
2. Write a neural network with 1 or more hidden layers
3. Solve the XOR
4. Understand how to build general classifiers

Lab 3 - Using PyTorch library to build multiple hidden layers

1. Understand how to use PyTorch to write a neural network
2. Write a neural network with multiple neural networks

Lab 4 - Adding convolutional layers to solve CNN (Convolutional Neural Network) problems

1. Write a convolutional network for the MNIST database