Fechada a lista de palestrantes para o 7º Behave Data Meetup:
Large-scale Machine Learning Pipeline | Pedro Calais e Thiago Akio
WorldSense uses large-scale machine learning to suggest linksto online publishers aiming to increase the revenue they make fromtheir content. In this talk we'll describe how we leverage Apache Spark and other popular open source tools to power our machine learning pipeline in a distributed cloud infrastructure, which include: how we train and build our models, how we monitor model quality over time, and how we've been experimenting with large-scale text-based deep learning models to suggest great links.
Pedro holds a doctorate degree from UFMG's Computer Science Department.His research interests are in machine learning, data mining, social networks andconnecting social sciences and social psychology theories to algorithms that process big data. As a software engineer in WorldSense, he is loving working in applied research in the industry.
Thiago holds an engineering degree in Control and Automation from UFMG's engineering school. Has researched about Fault Detection and Diagnosis in industrial plants using machine learning and statistical inference techniques, and is currently interested in machine learning for text/natural language processing for his master's degree. He is excited to be able to learn about and apply the state-of-the-art algorithms within WorldSense.
• "Algoritmos Ensemble com Árvores de Decisão" - Andressa Sivolella
• "Bancos de dados em grafos e Neo4j" - Péterson Sampaio Procópio Júnior
• "Excel e Data Science, Parte II" - Isaias Barroso
• "Modelo Matemático para Otimização de Rotas de Ônibus com Dados Coletados por IoT" - João Fábio
Detalhes sobre estas palestras podem ser vistos aqui (http://speakerfight.com/events/7o-behave-data-meetup/).