AI Student Ambassador Devinder Kumar: Applying Deep Learning in the Healthcare and Finance Sectors

Published: 02/07/2018  

Last Updated: 02/07/2018

The Intel® Student Ambassador Program was created to work collaboratively with students at innovative schools and universities doing great work in the Machine Learning and Artificial Intelligence space. I had the opportunity to get to know one of our student ambassadors, Devinder Kumar, and learn about his work with machine learning for the healthcare and finance industries.

Tell us about your background.
I am a first year PhD student at Vision and Image Processing (VIP) lab, University of Waterloo and Machine Learning Research Group (MLRG), University of Guelph. Before this, I completed my Masters in WAVE lab & VIP lab at University of Waterloo, as well as, a research engineer in LIP-6 - UPMC-Sorbonne University, Paris. My research centers on deep learning and its application in Computer Vision. Specifically, the research problem I am currently focusing on is: How to make current deep models interpretable and compact enough to be scalable for real-time client side applications.

What got you started in technology?
I always wanted to create something that could reach and help millions of people in their daily lives. I saw technology and especially machine learning as a tool that can help me in achieving this dream.

What projects are you working on now?
Currently I am working on two projects, one is in the domain of Explainable AI – where I develop approaches for explaining the decision-making process of deep neural networks. I explain this project more in an articles published on University of Waterloo – Engineering’s website and also on VICE Motherboard site. The other project that I just started is related to semi-supervised learning where the aim is to efficiently learn the association between input and output using very few examples. You can follow along with my work on my personal blog.

You recently won “Best Paper” award at Transparent and Interpretable Machine Learning (TIML) workshop at Neural Information Processing Systems (NIPS) 2017. Can you tell me about your project that led to this paper?
In the later part of 2016, when I was an intern at one of the biggest medical imaging device companies, I saw first-hand the need and demand to create explainable deep neural networks for the healthcare sector. My interest in this area grew with time and after my internship ended I started my PhD, with Explainable AI as my first project. During this project, we developed a new approach called Class-Enhanced Attentive Response (CLEAR) for visualizing the decision making process of deep neural networks. We applied CLEAR for visualizing the grading process of diabetic retinopathy diseases and wrote a paper on this which won the best paper award at the TIML workshop at NIPS 2017. Here is the link to the full version of the paper.

What trends do you see happening in technology in the near future?
I think the next big thing in technology would be the widespread use of AI in sectors which are largely considered "unconventional" for such technology like the energy or oil industries. AI for block-chain will be another big thing as well.

How are you planning to leverage artificial intelligence or deep learning technologies in your work?
I am using deep learning technologies to create better and more accurate predictive models in the field of medical imaging and finance. I am also working to make these models explainable as it is quite crucial for safety critical sectors to have machine learning models that explain their decisions.

What are you looking forward to doing with Intel?
I am planning to make the full use of the latest deep learning tools and computing platform generously provided by Intel to the community. I also plan to meet and potentially collaborate with other student ambassadors for exciting projects as well. By providing open source platforms and tools, including providing access to the Intel® AI DevCloud, Intel is doing a lot to help students in the machine learning community succeed.

What impact on the world do you see AI having? And do you see yourself as part of it?

As Andrew Ng said, "Artificial Intelligence is the New Electricity". As electricity transformed each aspect of our lives in the past for the better, AI will do the same. AI will enhance and augment the human experience. I see myself as playing a very small role as a creator of AI technologies that can help better day to day life of millions of people.

Want to learn more? Check out our Student Developer Zone, join Intel® Developer Mesh, or learn more about becoming a Student Ambassador

Interested in more information? Contact Niven Singh

Product and Performance Information


Performance varies by use, configuration and other factors. Learn more at