Himanshu Jain

PhD Student, Department of Computer Science and Engineering
Indian Institute of Technology, Delhi
email: himanshu.j689@gmail.com
CV


About Me

I am a senior PhD student in the Department of Computer Science at the Indian Institute of Technology, Delhi and a Google PhD Fellow. My PhD advisor is Dr. Manik Varma. I completed my bachelor's and master's degree in Mechanical Engineering from the Indian Institute of Technology, Kanpur in the year 2012.
My primary research interest is in Extreme Classification which is a new paradigm for solving ranking and recommendation problems. In particular, I have been focussing on two aspects -

  • Designing loss functions that give an unbiased estimate of the true loss even when evaluated on the incomplete ground truth data and naturally promote tail labels.
  • Developing extreme classification algorithms that work well with low dimensional deep features and scale to problems with hundreds of millions of labels.

I have applied my algorithms to solve real-world problems such as query-recommendation on a search engine, where they have shown significant improvements over existing systems.


Publications

  • H. Jain, V. Balasubramanian, B. Chunduri and M. Varma. Slice: Scalable linear extreme classifiers trained on 100 million labels for related searches. In Proceedings of the ACM International Conference on Web Search and Data Mining, Melbourne, Australia, Feburary 2019. [Best Paper Award]
    Bibtex source | Abstract | Download in pdf format | Supplmentary in pdf format | Code | Extreme Classification Repository | Blog

  • H. Jain, Y. Prabhu and M. Varma. Extreme Multi-label Loss Functions for Recommendation, Tagging, Ranking & Other Missing Label Applications. In Proceedings of the ACM SIGKDD Conference on Knowledge Discovery and Data Mining, San Francisco, California, August 2016.
    Bibtex source | Abstract | Download in pdf format | Supplmentary in pdf format | Code | Extreme Classification Repository

  • K. Bhatia, H. Jain, P. Kar, M. Varma and P. Jain. Sparse local embeddings for extreme multi-label classification. In Advances in Neural Information Processing Systems, Montreal, Canada, December 2015.
    Bibtex source | Abstract | Download in pdf format | Code | Extreme Classification Repository


  • Resources

    The Extreme Classification Repository

    This repository provides benchmark multi-label datasets and code that can be used for evaluating the performance of extreme multi-label algorithms. It also provides comparative results of various algorithms on all the benchmark datasets.


    Personal

    I love trekking, and aim to go for atleast 2 treks per year but as is usually the case, haven't been able to manage that :(
    So far I have been to the following treks:

    • 4 Mile trail + Panorama trail, Yosemite national park (2199 m)
    • Half Dome, Yosemite national park (2695 m)
    • Buran Ghati Trek (4572 m)
    • Brahmatal Trek (3733 m)
    • Chadar Trek to Lingshed (3900 m): Checkout the trek video
    • Pin Parvati Pass (5300 m): Checkout the trek video
    • Rupin Pass (4650 m)
    • Lamayuru - Hemis Shukpachan Trek (3500 m)
    • Sandakphu (3600 m)
    • Small treks around pune