Machine Learning & Data Science Projects

In 2017, I decided to have a change in career from being a maths teacher and Vice Principal. In that year, I studied Neural Networks, Machine Learning and Data Science. I completed a number of projects to put my knowledge into practice, some of which, were coursework for the Data Science specialization by Johns Hopkins University on Coursera. I'm pleased to be able to report that following this I was able to get a job with Canon Medical Research Europe in the AI Research group.

Randomized Hyperparameter Search for a Tensorflow CNN

A randomized hyperparameter search was carried out for a vanilla CNN network built in Tensorflow. The CNN was designed to classify images in the CIFAR-10 dataset.

UK MOT Test Results Analysis

This is an ongoing peoject to analyse the UK MOT test results for cars and small vehicles in 2016.

USA Tornados (1996-2010)

The leaflet package in R was used to create an interactive map of USA tornados.
(Data Science specialization Coursework)

Interactive plot of the Maunga Whau Volcano

The plotly package in R was used to create two interactive graphics.
(Data Science specialization Coursework)

Greedy, Risk and Reward.

A R Shiny app was created to demonstrate the risks and rewards in the simple dice game, called Greedy.
(Data Science specialization Coursework)

Weight Lifting Method Detection

Various Machine Learning Models were compared to find the best one suited to classify weight lifting methods.
(Data Science specialization Coursework)

Word Prediction Algorithm

A simple word prediction algorithm using n-grams was developed.
(Data Science specialization Coursework)