Vantage AI

Blog

Bayesian Optimization for quicker hyperparameter tuning

Finding a good set of hyperparameters with common techniques such as GridSearch can be time-consuming; especially when exploring large parameter spaces. Bayesian Optimisation can reduce the time spent on tuning—thus bringing better generalisation performance on the test set. This blog describes its mechanics and shows it in action.

Lees deze blog

Beating the Bookies with Machine Learning

Bookies, much like casinos, always set the odds in their favor. However, while the probabilities of card and dice games are known, darts results depend heavily on the players' skill. Can we leverage the power of machine learning to beat the bookie and make a profit?

Lees deze blog

Building a Machine Learning API in 5 minutes

Once you've successfully trained a machine learning model, you'll want to make it accessible so people can actually start using it. But how does that work? This post discusses how you can use Flask to build an API for your model in just five minutes.

Lees deze blog

Dealing with Large Datasets in Pandas

Data pre-processing and feature engineering are vital to successful completion of data science projects. However, these steps can involve a lot of waiting―especially as the size of your dataset increases. How can we optimize Pandas to make the most of available computing power?

Lees deze blog