I am currently a post-doctoral research associate at KU Leuven. I have completed my Ph.D. under the supervision of Prof. dr. Tias Guns. My research falls at the confluence of machine learning (ML) and combinatorial optimization problem (COP).

In my PhD, I have studied Decision-focused learning. In decision-focused learning, ML prediction is followed by COP for decision-making. The goal is to train the ML model, very often a neural network model, directly considering the error after the COP. The primary challenge in the implementation decision-focused learning is how to embed the COP into the ML training loop. To address this challenge, I have developed a differentiable optimizer, which enables passing the gradient through the COP for training the ML model. I am also interested in scalable decision-focused learning, so that it can be applied in real-life COPs, which are often NP-hard and time-consuming to solve.

Our survey article on Decision-Focused Learning for Predict-then-Optimize is available on https://arxiv.org/abs/2307.13565

• Jayanta Mandi, Victor Bucarey Lopez, Maxime Mulamba and Tias Guns. Decision-Focused Learning: Through the Lens of Learning to Rank. ICML, 2022, International Conference on Machine Learning, 2022 [paper] [Code] [Presentation] [Poster]

• Jayanta Mandi, Rocsildes Canoy, Victor Bucarey Lopez and Tias Guns. Data Driven VRP: A Neural Network Model to Learn Hidden Preferences for VRP. CP, 2021, International Conference on Principles and Practice of Constraint Programming, 2021 [paper] [Code] [Presentation]

• Maxime Mulamba, Jayanta Mandi, Michelangelo Diligenti, Michele Lombardi, Victor Bucarey Lopez and Tias Guns. Contrastive Losses and Solution Caching for Predict-and-Optimize. IJCAI, 2021, International Joint Conference on Artificial Intelligence, 2021 [paper] [Code] [Presentation]

• Jayanta Mandi and Tias Guns. Interior Point Solving for LP-based prediction+optimisation. NeurIPS, 2020, Advances in Neural Information Processing Systems, 2020 [paper] [Code] [Poster]

• Maxime Mulamba, Jayanta Mandi, Rocsildes Canoy, Tias Guns. Hybrid Classification and Reasoning for Image-based Constraint Solving. CPAIOR, 2020, 17th International Conference on the Integration of Constraint Programming, Artificial Intelligence, and Operations Research, 2020 [paper] [Presentation]

• Jayanta Mandi, Emir Demirović, Peter. J Stuckey and Tias Guns. Smart Predict-and-Optimize for Hard Combinatorial Optimization Problems. AAAI, 2020, AAAI Conference on Artificial Intelligence, 2020 [paper] [Poster]

• Dipankar Chakrabarti, Neelam Patodia, Udayan Bhattacharya, Indranil Mitra, Satyaki Roy, Jayanta Mandi, Nandini Roy, Prasun Nandy. Use of Artificial Intelligence to Analyse Risk in Legal Documents for a Better Decision Support. TENCON 2018, IEEE Region 10 Conference, 2018 [paper]

• Manisha Chakrabarty and Jayanta Mandi. Entropy-Based Consumption Diversity—The Case of India. Opportunities and Challenges in Development, Springer, Singapore, 2019. 519-540. [paper]

• Ashok Banerjee, Jayanta Mandi and Deepnarayan Mukherjee. Developing a comprehensive earnings management score (EMS). [article]

• Ideas for India. Jayanta Mandi, Manisha Chakrabarty and Subhankar Mukherjee. "How to ease Covid-19 lockdown? Forward guidance using a multi-dimensional vulnerability index". [article]

• Business Standrd. Ashok Banerjee, Jayanta Mandi and Deep N Mukherjee. "Earnings management in stressed firms". [article]