Study and analysis of the Kaggle Competitions to encourage the learning and the exchange of experiences of the machine learning community of Campinas

Resumo

Kaggle is the main platform of predictive and analytics models competition. The goal of this research project is to study and analyze the main Kaggle competitions, their datasets, and their winner teams solutions, thus, learning the best techniques and practices in the machine learning area. This information, summed up with Kaggle Learn courseware, was used as a subsidy for themes and activities of Kaggle Campinas Meetup meetings coordinated together with the research. The competition analysis produced an updated view of some of the bests techniques for each problem type, and the meetings helped to broaden the machine learning community of Campinas and Unicamp.