I try to get machines to make decisions by themselves. This amounts to doing basic and applied research on machine learning and computational intelligence with a team of researchers at Telenor Research and the Telenor-NTNU AI Lab.

Additionally, I advise a handful of startups based in Oslo Science Park on how the state-of-the-art in machine learning and computational intelligence can be adopted/adapted for their needs.

From an academic point of view, the general theme of my research in recent years has revolved around this question:

How can we create and/or manage computing entities, which naturally have limits to what they can know and do at any point in time, to get things done collectively?

Such entities interact with each other, learn and adapt to circumstances their interactions and individual limits lead them into. The behaviour of the collective is influenced by entities operating in this manner. Consequently, the collective may or may not get things done as desired. The need for understanding collective behaviour emerging via interactions, or even engineering desirable collective behaviour, has pulled me into research domains which go by the following names:

/ machine learning, specifically generalisation and reinforcement learning
/ multi-agent systems
/ computational economics
/ automated mechanism design
/ game theory
/ information theory
/ nature inspired computing

I have previously worked as a researcher (post-doc/research associate/phd student) primarily on the following projects:
Engineering Proprioception in Computing Systems at the University of Oslo
Adaptive Dynamic Ensemble Prediction Techniques at the University of Manchester
Market Based Control of Complex Computational Systems at the University of Birmingham

I also do guest lectures at the University of Oslo and NTNU.

A strong contender for my free time is being outdoors, sometimes at altitude.

arjun chandra

Email:
/ 'full.name'@telenor.com
/ 'name'@studix.com
/ 'surname'@ifi.uio.no

@boelger

View Arjun Chandra's profile on LinkedIn