Tirtharaj Dash

Assistant Professor (Grade-II), Dept. of CS & IS and APPCAIR, BITS Pilani, Goa Campus

prof_pic.jpg

D-168

BITS Pilani, Goa Campus

Zuarinagar, Goa 403726

I am working as an Assistant Professor (Grade-II) in the Department of Computer Science and also affiliated with Anuradha and Prashanth Palakurthi Centre for Artificial Intelligence Research (APPCAIR), BITS Pilani Goa Campus. My research areas of interest are Deep Learning, Neuro-Symbolic Models, Graph Representation Learning and Machine Learning for Science.

I have submitted my PhD thesis titled “Inclusion of Symbolic Domain-Knowledge into Deep Neural Networks” to the Department of Computer Science, BITS Pilani. I completed my doctoral research during Jan 2017 to April 2022 under the supervision of Senior Professor Ashwin Srinivasan. In my doctoral research, I focused on constructing deep neural networks from relational data and symbolic domain-knowledge—this resulted in beautiful combinations of neural computation with logical representation. The real-world applications of my doctoral research are in the broad area of drug discovery. Further, it has enormous potential to be adopted in many other real-world problems such as health science, social networks, robotics, etc. For now, I have made the abstract of my dissertation available here.

I received a Master of Technology (M.Tech) degree in Computer Science from VSSUT, Burla (one of the oldest engineering institutes in India) in the year 2014 and a Bachelor of Technology (B.Tech) degree in Information Technology from NIST Berhampur in the year 2012. I am a Silver Medalist in both M.Tech and B.Tech for my academic performances.

I have worked as an Assistant Professor in the School of Computer Science at NIST Berhampur for over a year during 2014 - 2015. I have also worked as IASc-INSA-NASI Summer Research Fellow at ISI Kolkata. I have qualified national-level competitive examinations such as GATE (twice) and UGC-NET.

Here is my semi-updated CV. References are available upon request.

latest news

Jun 23, 2022 Woo-hoo! Next phase in career starts at UCSD from Sep 2022.
Jun 1, 2022 Read our latest research on Compositional Relational Machines (CRMs).
May 13, 2022 PhD Journey ends! I submitted my PhD thesis. Read the abstract.
Apr 20, 2022 Our paper won the “Best Short Paper Award” at ACMSE-2022. Congratulations Gunjan, Abheesht and Harshit!
Mar 25, 2022 I am in the program committee for ICANN-2022.
Feb 22, 2022 Gunjan, Abheesht and Harshit got a paper accepted at ACMSE-22: [Preprint].
Feb 15, 2022 I will be talking AI at Davis Institute for AI on Feb 25, 2022.
Jan 16, 2022 Join my course CS F425:Deep Learning at Google Classroom.
Jan 1, 2022 Adieu, 2021. Wishing you all a very happy and healthy new year 2022.
Dec 31, 2021 I am in the program committee for IJCAI-ECAI 2022.

selected publications

  1. arXiv
    Composition of Relational Features with an Application to Explaining Black-Box Predictors
    Srinivasan, Ashwin, Baskar, A,  Dash, Tirtharaj, and Shah, Devanshu
    arXiv 2022
  2. SciRep
    A review of some techniques for inclusion of domain-knowledge into deep neural networks
    Dash, Tirtharaj, Chitlangia, Sharad, Ahuja, Aditya, and Srinivasan, Ashwin
    Scientific Reports Jan 2022
  3. ILP
    Using Domain-Knowledge to Assist Lead Discovery in Early-Stage Drug Design
    Dash, TirtharajSrinivasan, AshwinVig, Lovekesh, and Roy, Arijit
    In Inductive Logic Programming Jan 2022
  4. PNAS
    Transformational machine learning: Learning how to learn from many related scientific problems
    Olier, Ivan, Orhobor, Oghenejokpeme I.,  Dash, Tirtharaj, Davis, Andy M., Soldatova, Larisa N., Vanschoren, Joaquin, and King, Ross D.
    Proceedings of the National Academy of Sciences Jan 2021
  5. MLJ
    Inclusion of domain-knowledge into GNNs using mode-directed inverse entailment
    Dash, TirtharajSrinivasan, Ashwin, and Baskar, A.
    Machine Learning Nov 2021
  6. MLJ
    Incorporating symbolic domain knowledge into graph neural networks
    Dash, T.Srinivasan, A., and Vig, L.
    Machine Learning Jul 2021