Andre Mendes da Silva

Andre Mendes da Silva’s work involves using machine learning to create recommendation systems that can be applied to different fields. One direct application is the hyper agent (http://julian.togelius.com/Mendes2016HyperHeuristic.pdf) that uses artificial intelligence to select subagents to play general video game competitions. Another field is people analytics where he develops recommendation systems that are able to improve different selection processes, such as job applications, turnover rate, and career paths. In his research, he has developed a framework that uses sub-neural networks to collect features from different types of data that include categorical, numerical, video, and audio. These features are ultimately combined and used as input for a hyper-neural network that is able to predict the chance of an applicant becoming a top manager or the chance that a talented employee will leave a company. While at NYU Shanghai da Silva plans to develop research and work with the renowned people in his field, such as professors Nan Cao, Zheng Zhang, and Keith Ross.