Machine Learning & Deep Learning Portfolio

Nur A Jaman — Academic/Research Portfolio

Abstract

This portfolio presents my systematic exploration of Machine Learning and Deep Learning, spanning classical algorithms, dimensionality reduction, neural networks, computer vision, natural language processing, and applied AI.

Each notebook is structured in the spirit of a research paper — with clear motivation, methods, and results — forming both a learning journey and a portfolio of applied research efforts.

Beyond experimental ML/DL projects, this portfolio also highlights production systems and professional projects, including Science Master BD, a paid EdTech platform deployed under real-world constraints. This dual focus demonstrates my ability to bridge academic innovation with practical, scalable system design.

I. Classical Machine Learning Algorithms

II. Neural Networks

III. Computer Vision

IV. Sequential Models

V. Natural Language Processing

VI. Applied AI Projects

VII. Production Systems & Professional Projects
  • Research Buddy V1.1 — White Paper 📄

    A modular ML/AI system for research paper analysis, integrating classification, keyword extraction, summarization, and user libraries. Includes a future development roadmap with conversational LLMs, RAG, and an AI Agent mode.

  • ScienceMaster — EdTech Platform Architecture (White Paper) 💼

    A production-grade EdTech platform built with MERN, AWS EC2/Lightsail, and MongoDB replica sets. Developed under real-world constraints (no DevOps hires, budget limits). This is a paid project with live users — demonstrating the transition from research to scalable deployment.