Probabilistic Graphical Models (PGMs) Research Internship

About Internship

Assistant Professor from Bhagwan Parshuram Institute of Technology, India, is accepting candidates interested in the field of Computer Science Engineering, in the research area of Artificial intelligence and Machine learning.

This advanced research position focuses on developing a unified computational framework that merges graph theory with statistical mechanics and probability theory. By utilizing Probabilistic Graphical Models (PGMs), the project aims to encode structural dependencies within highly uncertain environments. The researcher will work extensively with Bayesian Networks (Directed Acyclic Graphs) for causal reasoning and decision-making, alongside Markov Random Fields (MRFs) (Undirected Graphs) to model spatial and contextual joint probabilities. This role is highly mathematical, bridging rigorous statistical theory with concrete computational implementations. Roles & Responsibilities: 1. Integrate structural graph theory with joint probability distributions to model complex, multi-variable dependencies in high-dimensional data spaces. 2. Design and implement Directed Acyclic Graph (DAG) structures to facilitate exact and approximate probabilistic inference, structure learning, and structural causal reasoning. 3. Construct undirected graphical structures (MRFs) to map cyclic, spatial, and multi-directional contextual dependencies, optimizing joint distributions via energy-based formulations. 4. Develop, evaluate, and benchmark message-passing algorithms and approximate inference methods . 5. Test framework performance on messy, real-world high-dimensional datasets, ensuring the models maintain a high balance between computational efficiency and semantic interpretability. 6. Translate mathematical derivations and experimental codebases into structured technical reports, algorithmic proofs, and draft papers targeting elite AI/ML publications.