statistics

Using the GPU for gradient descent boosting

Running machine learning algorithms on large amounts of data can take considerable time. There are multiple ways of speeding up your code. The most obvious is to properly parallize your code and/or - assuming R or python is used - replace certain functions that cause a bottleneck to faster languages such as C++ or Julia.

Testing multilevel Bayesian models with ordered categorical predictors

Why argue for Bayesian models? Most researchers and data scientists have specific - domain - knowledge about the subject they analyse data for. In a Bayesian analysis framework this knowledge can be refereed to as Prior and the effect und uncertainty surrounding this.