I am currently a Ph.D. candidate at the University of Illinois at Chicago, Department of Information and Decision Sciences. Here at UIC, I am working with Prof. Nadarajah. My research focuses on developing and applying machine learning, statistical, and optimization algorithms to enhance the accuracy of prediction and decision making in applications such as user behavior modeling, portfolio management, and investment.
I enjoy problem-solving and love coding. Ever since writing my first algorithm in Python, I have been obsessed with the idea of using programming languages to solve practical problems. I also enjoy building a statistical and mathematical frame for modeling problems as it challenges so many aspects of my intellect: creativity, sequential thinking, and problem-solving. Finally, I like to be challenged by learning new materials quickly.
- Our paper "Interpretable User Models via Decision-rule Gaussian Processes" got accepted in Advances in Approximate Bayesian Inference (NuerIPS 2019).
- Meeting Corporate Renewbale power targets, Under review in Management Science (with S. Nadarajah, A. Trivella) [PDF]
In this paper, we design procurement portfolios for companies which have committed to procuring a percentage of future power demand from renewable sources. Computing multi-stage procurement portfolios for these companies results in intractable Markov decision processes. Using methodologies in approximate dynamic programming, we develop a novel and near-optimal heuristic for decision making on realistic instances.
- Interpretable User Models via Decision-rule Gaussian Processes, NuerIPS 2019 (AABI symposium) (with S. Nadarajah, T. Tulabandhula)
Models of user behavior are critical inputs in many prescriptive settings and can be viewed as decision rules that transform state information available to the user into actions. Gaussian processes (GPs), as well as nonlinear extensions thereof, provide a flexible framework to learn user models in conjunction with approximate Bayesian inference. However, the resulting models may not be interpretable in general. We propose decision-rule GPs (DRGPs) that apply GPs in a transformed space defined by decision rules that have immediate interpretability to practitioners. We illustrate this modeling tool on a real application and show that structural variational inference techniques can be used with DRGPs. We find that DRGPs outperform the direct use of GPs in terms of both out-of-sample performance and the quality of optimized decisions. These performance advantages continue to hold when DRGPs are combined with transfer learning.
- Investment under Limited Long-Term Information, Working Paper (with S. Nadarajah, A. Kelevin, S-E. Feleten )
The limited availability of long-term information about the uncertainties for decision-making problems (i.e., investment in Hydropower plant capacity) increases the potential impact of model misspecification and often lead to suboptimal decisions. To overcome this issue, we propose an algorithm that leverages statistical information in the medium-term and handles the lack of long-term information using robust optimization.
- Overbooking in Network of Storage Assets, Technical report (with S. Nadarajah, T. Tulabandhula )
SKILLS AND BACKGROUND
- Machine Learning
- Deep Learning
- Reinforcement Learning
- Applied statistics
- Python (Pandas, Pytorch, Tensorflow, Keras, Scikit-Learn, Scipy)
- R, MATLAB
- SQL, Spark
- Unix, AWS
PhD in Information and Decision Sciences, University of Illinois at Chicago(Exp. 2020)
BSc in Industrial Engineering, AmirKabir University, 2015
Graduate Reaserch Assistant in Artificial Intelligence
- Designed and implemented interpretable user models to predict users’ behavior using inference and transfer learning; tested the algorithm on corporate data; increased the accuracy of overbooking decisions by 10%.
- Developed and implemented near-optimal stochastic decision-making algorithms to decrease the cost of constructing a portfolio of a financial commodity by 4%.
- Proposed an algorithm for decision making under limited long-term information of uncertainties; calibrated a statistical model on time series corporate data.
Graduate Teaching Assistant in Data science
- Machine learning: Deep learning, bayesian inference, reinforcement learning
- Foundation of optimization: Linear and integer programming
- Programming: Provided tutorials in Pytorch
- Projects and leadership: Mentored graduate students in deep learning projects such as object detection and image classification (including CNN and RNN models).
Data science Mentor
- Cloudbakers: Led a group of graduate students in gathering data and creating a pipeline for clustering repositories in GitHub to evaluate their health. Presented results to stakeholders, Fall 2019
- Varuna (Startup): Collaborated with co-founders and graduate students on processing data and building a learning algorithm to predict failures in water purification plants , Spring 2019
|Interpretable User Models via Decision-rule Gaussian Processes INFORMSAnnual Meeting, Seattle, Washington||OCT 2019|
|Meeting Corporate Renewable Power Targets INFORMSAnnual Meeting, Seattle, Washington||OCT 2019|
|Overbooking in Network of Storage Assets Production and Operations Management Society Annual Conference, Washington D.C. (POMS)||May 2019|
|Meeting Corporate Renewable Power Targets Production and Operations Management Society Annual Conference (POMS)||May 2019|
|Investment under Limited Long-Term Information Production and Operations Management Society Annual Conference (POMS)||May 2019|
|Dual Reoptimization based Approximate Dynamic Programming INFORMS Annual Meeting, Phoenix, Arizona||Nov 2019|
|Meeting Corporate Renewable Power Targets Production and Operations Management Society Annual Conference, Houston, Texas (POMS)||May 2019|
|Meeting Corporate Renewable Power Targets Manufacturing & Service Operations Management Annual Conference, Dallas, Texas (MSOM)||Jul 2018|