Grouping Parliamentary Constituencies using Clustering Analysis
Do the London, Scotland, or the Red Wall parliamentary constituencies form cohesive groups? We explore using different choices of clustering techniques.
Graph Neural Networks for link predictions
Or yet in wise old Ravenclaw, If you have a ready mind, Where those of wit and learning, Will always find their kind.
Store Sales - Time Series Forecasting Challenge
We enter a Kaggle competition and use time-series forecasting to predict store sales on data from Corporación Favorita, a large Ecuadorian-based grocery retailer.
One-Hit Wonders
According to Wikipedia, a one-hit wonder is any entity that achieves mainstream popularity, often for only one piece of work, and becomes known among the general public solely for that momentary success. In this project we use the publicly available data from Spotify to find artists in the music industry who can be regarded as one-hit wonders.
Oh Hell! - Reinforcement Learning for solving card games
In this article we are going to plays cards. Well, not exactly ... with the power of reinforcement learning we are going to train intelligent agents that will play the game for us!
Reinventing Test and Trace: A Bayesian Approach For Modelling SARS-CoV-2 Setting-Specific Transmission
One of the key failures many countries had in dealing with the COVID-19 pandemic was a lack of coherent, data-driven analysis on setting-specific transmission risk. In this article, we outline a novel first-principles hierarchical Bayesian model for setting-specific transmission that, alongside the corresponding data collection methodology, could help governments in determining pandemic control policy
Predicting aqueous solubility with neural networks
Predicting aqueous of chemical compounds without the presence of a physical sample would make the process of drug discovery easier. Can machine learning help us achieve this?
Random Coin Flips
Humans are inherently bad at understanding randomness and probability. The classic example is the coin-flip. In this post use statistical hypothesis testing to show this phenomenon.
Computational Approach to Evolutionary Game Theory
From abstraction to simulation, we try to solve the same evolutionary problem using both analytical and computational techniques but do we reach the same conclusions?
Visualising the Intersection of Disease and the Human Genome
Since the Human Genome Project completed its survey of Homo sapian genes in 2003, we have had access to an incredibly powerful dataset for understanding disease and discovering new diseases. In this post, we demonstrate an interactive web app visualising such data.