Event Diary - The Fifth Elephant 2018: Neural Field-aware Factorization Machines in Mobile Advertising

Posted on September 05, 2018
By Team InMobi

InMobi regularly hosts, participates in and contributes to a series of events, meetups, webinars, sharing best practices with partners and thought leaders, across the globe. In this edition of the InMobi Event Diary, we are looking at our recent visit to one of India’s top business schools - the Indian Institute of Management in Bangalore, where we addressed students of the Big Data and Analytics Certificate Programme.

In this edition of the event diary, we cover our presence at the famed data science conference - Fifth Elephant 2018. Anthill Inside+Fifth Elephant is one of India’s most renowned data science conferences and this year, the conference was held at the Bangalore NIMHANS Convention center.

The conference delves in the areas of Natural Language Processing (NLP), Reinforced Learning (RL), Deep Learning (DL) and Big data. The conference aims to bridge the gap between research in these areas and their usage/application in the industry. With the recent advancements in hardware efficiency, a lot the data-science research ideas are now successfully being ported to reality for specific use cases.

We at InMobi are also trying to bridge the same gap with newer deep learning models for a bunch of problems in the mobile advertising and marketing domain. At Anthill Inside, The Fifth Elephant’s adjunct event, we presented one such model: Neural Field aware Factorization Machine (NFFM). The NFFM is a model that predicts user behaviour such as install rate, video completion rate or click through rate, when engaging with mobile ads.

NFFM combines the power of field-aware factorization machines which can easily reduce the dimensionality of high cardinality features and find the second order interactions across features. NFFM then runs these newly discovered second order interactions through neural networks to find higher order interactions and share the required behaviour prediction.

Watch the complete video to learn more about the NFFM model and its application in mobile advertising.

Or, view the complete deck here.

Subscribe to the InMobi ML-AI newsletter to stay up-to-date on the latest in the field of data sciences from us.