phd-portfolio

An ultra-minimalist, Jekyll-based portfolio template for academics.

Complete scRNA-seq analysis pipeline for 10x Genomics PBMC 3k dataset using Python and Scanpy.

Research-level implementation of unsupervised anomaly detection using KMeans, DBSCAN, Isolation Forest, and deep Autoencoders. Applied to IoT sensors, financial fraud, network intrusion, and time-series fault detection. Built for PhD-oriented ML portfolios.

Hybrid BM25 and dense-retrieval investor recommender with reciprocal-rank fusion, evaluation and LLM-ready reranking.

A research-style Python project for geographic question answering: natural language questions are mapped to explicit semantic representations and executed over a geographic knowledge graph, with rule-based parsing, evaluation metrics, and error analysis.

Research-oriented natural-language data analysis system with intent detection, guarded execution and reproducible evaluation.

Research-grade reinforcement learning framework for robot navigation, covering discrete, obstacle-aware, continuous-control, and multi-agent environments with PPO and DQN, full evaluation pipeline, reproducible experiments, and LaTeX paper template for PhD-level research.

Reproducible Earth Observation pipeline for multi-temporal forest change detection and uncertainty-aware evaluation.

Explainable transfer-learning framework for plant disease classification and model reliability analysis.

Research portfolio connecting my work on multimodal learning, retrieval systems, contrastive learning, embedding geometry, and AI evaluation.

Python simulation of BPSK, OFDM, and Alamouti STBC — physical layer wireless communications from scratch

Explainable deep-learning research pipeline for dental cavity detection from smartphone images.

Reproducible FAIR-inspired Monte Carlo modeling for quantitative cybersecurity risk analysis in cloud environments.

Personal portfolio of Dr. Mallikarjuna Thippana — computational biologist specializing in genomics & epigenomics (bulk RNA-seq, ChIP-seq, ATAC-seq, CUT&RUN), single-cell & spatial multi-omics, and statistical ML for biological data.

Open-source ML pipeline for Pueraria isoflavone bioactivity prediction. v0.6.1: 5-target SAR (ER-α/β, PI3K, AKT1, MMP-9), CV AUC 0.91-0.95. PhD application portfolio 2027.