Machine Learning is Not Just Statistics

We are increasingly hearing that ML is just a "sexy" name for old statistical techniques. This article aims to prove that this is not the case.

On one hand, there are enthusiasts who believe that by using the “keras” library, they will immediately outpace their competition. On the other hand, there are those who see deep learning as an inflated buzzword and predict the arrival of an impending AI winter.

The author of the article counters the view that ML is merely another name for statistics. Statistics is a field of mathematics, while machine learning is a group of computational algorithms from the realm of computer science. In the case of reinforcement learning, working with a dataset can be based on principles that are significantly different from those in predictive modelling. One can notice many successful individuals in the field of machine learning who do not excel in mathematics.

“When you start a startup, you call it AI; when you’re looking for people, you talk about ML; and when you’re developing it, it’s logistic regression.”
Everyone on Twitter

Source: https://towardsdatascience.com/no-machine-learning-is-not-j…