Have you heard about Keras but never known what is it about or where to start with? In this post we will explain Keras basics and we will use it to build a handwritten digit classifier with a high ...

Some machine learning problems involve classifying an object into one of N classes. These are called multiclass classification problems, as opposed to binary classification, where there is just a p...

They are on everyone’s lips: every single post today seems to talk about deep neural networks and the bewildering variety of applications they are used for. Speech recognition, computer vision, nat...

In this post I will talk about the underfitting and overfitting phenomena, and how model complexity affects them. I will also explain the bias-variance trade-off. We will demonstrate these concept...

In this post I will talk about one of the most basic models in Machine Learning: logistic regression. This post doesn’t assume any previous knowledge of logistic regression. Some prior knowledge on...