18:00 - 18:30 : Gathering
18:30 - 20:00 : The Race for AI: Lessons learned while building deep learning architectures for recommendation systems.
Deep Learning models have been gaining increasing attention in the recommendation systems community, replacing some of the traditional methods. The sparse nature of the problems in this domain and the different types of inputs offer unique challenges for feature engineering and architecture planning, in order to balance between memorization and generalization.
During the past year the algorithms team in Taboola moved all of our algorithms to deep learning models and in this talk we will share the lessons we learned doing so. Specifically, we will talk about building neural networks with multiple input types (click history, text and pictures); feature engineering in the deep learning era; capturing interactions between features using both deep dense architectures and Factorization Machine models; Tradeoffs between deep models, shallow models and the combination of the two; and other tips regarding network architectures.
Ofer Alper is an Algorithm Team Leader at Taboola, the world's largest discovery platform. In his current position he is responsible for building a machine learning based recommendation system, using neural network, deep learning and matrix factorization. Before joining Taboola Ofer worked in the Algo-Trading industry developing algorithms and data science models. Prior to that he was the R&D Manager of Mocospace, one of the largest mobile social networks in the US. Ofer has over 17 years of experience in many areas including software engineering, algorithm development, data science and machine learning.