Hello World!

Welcome to MLG@PUT!

Posted by Marek Wydmuch on 5th Apr 2018

Table of Contents

About us and this site

Machine Learning Group @ Poznań University of Technology was founded in 2016 by Krzysztof Dembczyński Ph.D. The aim of our group is to bring together students and researchers interested in fundamental aspects of machine learning and artificial learning. Every week we present, analyze and discuss one topic using only whiteboard and markers to understand in depth the theory behind machine learning algorithms.

This website is our new initiative and aims to serve as a backlog of our meetings.

See About subpage for more information.


This part is for my colleagues to show them that Latex syntax works flawlessly.

Decomposition of the expected loss over all examples (risk):

$$ \begin{align} L_\ell(h) &= \mathbb{E}_{(x,y) \sim P}[\ell(y, h(x))] \\ & = \int_{X \times Y} \ell(y, h(x))P(x, y)dxdy \\ & = \int_{X} \left( \int_{Y} \ell(y, h(x))P(y|x)dy \right) P(x)dx \\ & = \mathbb{E}_x[L_\ell(h|x)] \\ \end{align}$$

$L_\ell(h|x)$ is the conditional risk of $\hat{y} = h(x)$ at $x$.