Avatar

Sanket Shah

Research Engineer

Singapore Management University

I am a Research Engineer at Singapore Management University (SMU) advised by Prof. Pradeep Varakantham. My current work focuses on using Reinforcement Learning to solve real-world problems in Transportation and Security. I am more broadly interested in understanding how to use my knowledge of AI to be of use to society.

Prior to this, I spent a year at Microsoft Research India during which I worked on Natural Language Processing with Dr. Sundararajan Sellamanickam and Information and Communications Technology for Development with Dr. Colin Scott and Dr. Bill Thies.

Interests

  • AI for Social Good
  • Reinforcement Learning
  • Machine Learning

Education

  • B.E. (Hons.) in Computer Science, 2017

    Birla Institute of Technology and Science, Pilani

Experience

 
 
 
 
 

Research Engineer

Singapore Management University

Nov 2018 – Present Singapore
Authored two first-author research papers that use Reinforcement Learning (RL) to address Sequential Decision Making problems that underlie societal challenges in Transportation and Security.
 
 
 
 
 

Research Intern

Microsoft Research India

Jan 2017 – Jun 2017 Bangalore, India

• Helped build an Android app that aimed to augment local peer-to-peer file transfer like Bluetooth (a substitute to the internet for media acquisition in low resource communities) by creating a barter economy.

• Helped pilot the application in a village in Bihar, India along with my advisor and local partners from the region.

 
 
 
 
 

Research Intern

Microsoft Research India

Dec 2016 – Aug 2016 Bangalore, India
Investigated the ‘explainability’ of Recurrent Neural Networks in terms of compositional linguistic structures like ‘and’ and ‘but’ for the task of Sentiment Analysis in English.
 
 
 
 
 

Research Intern

Philips India Ltd.

May 2016 – Jul 2016 Bangalore, India

• Prototyped the conversation engine for a wearable device to assist the elderly.

• Helped design an annotation scheme for patient medical records.

 
 
 
 
 

Research Intern

National Centre for Polar and Ocean Research

May 2015 – Jul 2015 Goa, India
Performed pixel-based supervised and unsupervised learning on hyper-spectral satellite imagery to study the spectral characteristics of supraglacial lakes in the Antarctic.

Recent Publications

Neural Approximate Dynamic Programming for On-Demand Ride-Pooling

We add future information to ride-pooling assignments by using a novel extension to Approximate Dynamic Programming.

Solving Online Threat Screening Games using Constrained Action Space Reinforcement Learning

We reformulate Threat Screening Games, a kind of Stackelberg Security Game, as an MDP with constraints on the action space.

Recent Posts

My Ride-pooling Journey

A blog post about the joyride my paper ‘Neural Approximate Dynamic Programming for On-Demand Ride-Pooling’ took me on.