Project Overview

1 - Iris Classifier

Introduction to ML classification, using SVMs, Decision Trees, and k-Nearest Neighbours.

Models: SVM, Decision Tree Classifier, kNN Classifier          Dataset: SKLearn Iris

2 - Titanic Survival

Predicting the survival of passengers on the Titanic using feature engineering.

Models: SVM, Decision Tree Classifier          Dataset: Kaggle Titanic Survival

3 - Linear & Logistic Regression

Constructing multi-dimensional linear and logistic regressors from scratch and testing their performance.

Models: Linear Regressor, Logistic Regressor          Dataset: Fish Market

4 - House Price Prediction

Predicting house prices from a large number of features using Linear Regression and Random Forest Regression.

Models: Linear Regressor, XGBoost, Random Forest Regressor          Dataset: Kaggle House Prices (Iowa)

5 - Hand-Drawn Digits Classification

Classifying hand-drawn digits using SVM and expanding the dataset using Data Augmentation.

Models: SVM          Dataset: SKLearn Digits

6 - Multi-Layer Perceptron

Constructing a multi-layer perceptron from scratch and testing its performance.

Models: Multi-Layer Perceptron          Dataset: Fish Market

7 - Boiling Point Prediction

Prediction the boiling points of organic molecules based on descriptors and fingerprints extracted from their structure.

Models: XGBoost, Random Forest Regressor          Dataset: Boiling Points

8 - Bayesian Optimisation

Optimising experimental conditions with multiple objectives and constraints using bayesian optimisation.

Models: Bayesian Optimisation

9 - Playing Card Image Classification

Classifying a large image dataset of playing cards using a Convolutional Neural Network.

Models: Convolutional Neural Network          Dataset: Kaggle Playing Card Images