Challenges in Representation Learning: A report on three machine learning contests

TL;DR

This paper reports on three machine learning contests from ICML 2013, highlighting datasets, results, and insights into challenges in representation learning across different domains.

Contribution

It introduces datasets and summarizes results from three distinct challenges, offering guidance for future contest organization and knowledge extraction.

Findings

Datasets for black box, facial expression, and multimodal learning challenges created.

Summarized competition results and insights into challenge difficulties.

Provided recommendations for future challenge design and knowledge gain.

Abstract

The ICML 2013 Workshop on Challenges in Representation Learning focused on three challenges: the black box learning challenge, the facial expression recognition challenge, and the multimodal learning challenge. We describe the datasets created for these challenges and summarize the results of the competitions. We provide suggestions for organizers of future challenges and some comments on what kind of knowledge can be gained from machine learning competitions.