Talks & Tutorials

Ayan Acharya

Senior Data Scientist
Netflix Inc.
Google Scholar Page: [Link]
Resume: [PDF]

About Me

I am working as a senior data scientist at Netflix Inc. If you love ML, NLP and are interested in applying your knowledge in solving challenging real-world problems associated with health-care, retail services, text mining, image, and audio analytics, do drop me an email and I would be happy to collaborate.
I graduated from Department of Electrical and Computer Engineering at University of Texas at Austin, USA in July 2015, where I was associated with Intelligent Data Exploration and Analysis Laboratory (IDEAL) led by Dr. Joydeep Ghosh and Machine Learning Research Group led by Dr. Raymond J. Mooney. In the past, I have also collaborated with Dr. Mingyuan Zhou from McCombs School of Business and with people from Applied Research Laboratories. My research focus as a graduate student was on developing generative models and efficient inference algorithms for solving transfer learning problems prevalent in text document analysis, social network study, recommender systems, object recognition from images and evolution of financial data. As an intern, I have worked in places like eBay Research Lab, Qualcomm Inc and Yahoo! Research, Sunnyvale, CA and Cognitive Scale, Austin, TX. Before joining UT Austin, I completed my Bachelor's degree in Electronics and Telecommunication Engineering from Jadavpur University, Kolkata, India in the year 2009.


Research Interests

Research Summary

In several applications, scarcity of labeled data is a challenging problem that hinders the predictive capabilities of machine learning algorithms. Additionally, the distribution of the data changes over time rendering models trained with older data less capable of discovering useful structure from the newly available data. Transfer learning is a convenient framework to overcome such problems where the learning of a model specific to a domain can benefit the learning of other models in other domains through either simultaneous training of domains or systematic sequential transfer of knowledge from one domain to the others. In my thesis, in all the approaches related to simultaneous learning, a low dimensional space is maintained that is shared across multiple domains and transfer of information may take place via such shared space. For sequential knowledge transfer, parameters of the model trained with data from an older domain are carefully adapted to fit the new distributions. Applications of such frameworks in problems like text classification, object recognition from images, network modeling for community detection and evolution of count data have shown promising results so far. Simultaneous knowledge transfer has also been integrated with active learning to gain additional benefits in domains where labeled data is expensive to obtain. In my research, numerous probabilistic low-rank factorization models (both parametric and non-parametric) are used which help capture dependencies across multiple related domains in a shared low dimensional space. A significant amount of research effort is also directed towards the development of efficient inference mechanisms for the proposed models. Please check the Research page for further details.

Scholarly Service Activities

Find Me