Problems I solved: Reinforcement learning, Pricing, Forecasting, RankingOptimization I addressed: Constrained, High dimensional, Multi-objectiveML models I used: Classical ML models, Deep Neural networks, Large Language Model (LLM)

Supervised learning: Recommendation, RankingUnsupervised learning: Generative modeling, Limit order book modelingSemi-supervised learning: Anomaly detectionAutomatic machine learning (AutoML)

Machine learningPredictive analyticGenerative model

I worked on models and algorithms in deep learning at Information Initiative @Duke (iiD), Duke University.

Segment, classify and automatically detect anomalies in land and sea-based hyperspectral surveillance data;Develop a fast probabilistic fusion of the multiple wavelength images of the same scene, incorporating segmentation, classification, and anomaly detection as latent variables in a joint Bayesian inference scheme;Combine Markov chain Monte Carlo with deterministic methods such as variational approaches and other fast optimizers.