Shameek Mandal

Academic Projects

Mon, 01 Jan 0001 00:00:00 +0000

Application of Power K Means for Multiview Data | Master’s Final Year Project

Formulated a novel clustering framework - Cluster Weighted Kernel Power K Means extending Multiview Kernel Power K-Means to optimize non-convex objective functions and mitigate local minima limits.

Framework: Kernelized data projections via Reproducing Kernel Hilbert Space (RKHS).
Optimization: Solved non-convexity constraints utilizing the Majorization-Minimization (MM) algorithm to dynamically weight inner clusters across multiple views.
Benchmarks: Validated clustering accuracy on simulated frameworks and Caltech 101 dataset using NMI and ARI metrics.

Project Link

Academic Projects

Mon, 01 Jan 0001 00:00:00 +0000

Application of Power K Means for Multiview Data | Master’s Final Year Project | Supervisor - Prof. Swagatam Das, ISI Kolkata

Formulated a novel clustering framework extending Multiview Kernel Power K-Means to optimize non-convex objective functions and mitigate local minima limits.

Framework: Kernelized data projections via Reproducing Kernel Hilbert Space (RKHS).
Optimization: Solved non-convexity constraints utilizing the Majorization-Minimization (MM) algorithm to dynamically weight inner clusters across multiple views.
Benchmarks: Validated clustering accuracy on simulated frameworks and Caltech 101 using NMI and ARI metrics.

Journey

Mon, 01 Jan 0001 00:00:00 +0000

[Aug 2026 - ] | Pennsylvania State University

Role: PhD Student & Graduate Teaching Assistant, Department of Statistics
Focus: To work on my courseworks, and deep diving into Theoretical Machine Learning.

[Dec 2025 - May 2026] | Prefr (CRED)

Role: Data Scientist
Core Tasks: Designed statistical validation frameworks to evaluate the conceptual soundness and stability of Credit Risk Machine Learning models under regulatory constraints.

[July 2022 - Sept 2025] | DMI Finance Private Limited

Role: Credit Risk Associate
Core Tasks: Formulated regulatory statistical validation architectures to evaluate and stress-test high-volume customer acquisition credit risk ML models and built statistical segment based decision rules for customer acquisition strategy.

[July 2017 - June 2022] | Indian Statistical Institute, Kolkata

Degree: Completed Bachelors and Masters degree on foundational studies in advanced mathematical statistics. Master’s specialisation - Computational Statistics

Shameek Mandal

Mon, 01 Jan 0001 00:00:00 +0000

Incoming PhD Student | Department of Statistics, Pennsylvania State University, PA, USA
B.Stat and M.Stat | Indian Statistical Institute (ISI), Kolkata, India

Welcome to my digital terminal. I am a Statistics scholar transitioning from the industry to academia. Previously, I worked as a Credit Risk Associate / Data Scientist developing and regulating robust financial statistical models for customer acquisitions in the Indian Fintech sector.

Interests:

Statistical Machine Learning - Distribution Shifts, Data Mining, Class Imbalance
Non Parametric and Robust Statistics
Real Madrid C.F.
Tac FPS, Neo Noir Classics