User profiles for Abir De
![]() | Abir DeAssistant Professor, CSE, IIT Bombay Verified email at cse.iitb.ac.in Cited by 1788 |
Learning and forecasting opinion dynamics in social networks
Social media and social networking sites have become a global pinboard for exposition and
discussion of news, topics, and ideas, where social media users often update their opinions …
discussion of news, topics, and ideas, where social media users often update their opinions …
Discriminative link prediction using local, community, and global signals
A De, S Bhattacharya, S Sarkar… - … on Knowledge and …, 2016 - ieeexplore.ieee.org
Predicting plausible links that may emerge between pairs of nodes is an important task in
social network analysis, with over a decade of active research. Here, we propose a novel …
social network analysis, with over a decade of active research. Here, we propose a novel …
Grad-match: Gradient matching based data subset selection for efficient deep model training
The great success of modern machine learning models on large datasets is contingent on
extensive computational resources with high financial and environmental costs. One way to …
extensive computational resources with high financial and environmental costs. One way to …
Nevae: A deep generative model for molecular graphs
… Moreover, in contrast with the state of the art, our decoder is able to provide the spatial …
This work was done when Abir De was a post doctoral researcher at MPI-SWS, Germany. …
This work was done when Abir De was a post doctoral researcher at MPI-SWS, Germany. …
Enhancing human learning via spaced repetition optimization
Spaced repetition is a technique for efficient memorization which uses repeated review of
content following a schedule determined by a spaced repetition algorithm to improve long-term …
content following a schedule determined by a spaced repetition algorithm to improve long-term …
Regression under human assistance
Decisions are increasingly taken by both humans and machine learning models. However,
machine learning models are currently trained for full automation—they are not aware that …
machine learning models are currently trained for full automation—they are not aware that …
Classification under human assistance
Most supervised learning models are trained for full automation. However, their predictions
are sometimes worse than those by human experts on some specific instances. Motivated by …
are sometimes worse than those by human experts on some specific instances. Motivated by …
Learning a linear influence model from transient opinion dynamics
Many social networks are characterized by actors (nodes) holding quantitative opinions
about movies, songs, sports, people, colleges, politicians, and so on. These opinions are …
about movies, songs, sports, people, colleges, politicians, and so on. These opinions are …
Deep reinforcement learning of marked temporal point processes
U Upadhyay, A De… - Advances in neural …, 2018 - proceedings.neurips.cc
In a wide variety of applications, humans interact with a complex environment by means of
asynchronous stochastic discrete events in continuous time. Can we design online …
asynchronous stochastic discrete events in continuous time. Can we design online …
Differentiable learning under triage
Multiple lines of evidence suggest that predictive models may benefit from algorithmic triage.
Under algorithmic triage, a predictive model does not predict all instances but instead …
Under algorithmic triage, a predictive model does not predict all instances but instead …