Skip to content

README.md

Latest commit

 

History

History
47 lines (31 loc) · 2.35 KB

README.md

File metadata and controls

47 lines (31 loc) · 2.35 KB
Raw

MIT Applied Data Science

Portfolio notes and coursework from MIT Professional Education: Applied Data Science - Leveraging AI for Effective Decision-Making.

This repository documents applied machine learning, statistical reasoning, data visualization, supervised and unsupervised learning, neural networks, recommendation systems, time-series analysis, and computer vision practice completed through the program.

Why This Repository Matters

My main professional focus is bioinformatics and computational biology. This program strengthened the AI/ML layer of that work: how to frame prediction problems, evaluate models, avoid overclaiming, and communicate model behavior clearly to scientific and business stakeholders.

Biomedical Imaging Capstone

The capstone project focused on computer vision for biomedical imaging, using microscopy/RBC images for malaria detection.

Highlights:

  • Built preprocessing workflows with OpenCV.
  • Implemented CNN architectures with TensorFlow/Keras.
  • Compared custom CNN performance against transfer-learning baselines.
  • Achieved 97% accuracy and 99% sensitivity in the project setting.
  • Emphasized sensitivity because missed positive cases are especially costly in screening contexts.

Skills Demonstrated

Area Examples
Programming Python, Jupyter notebooks, pandas, NumPy
Machine learning Regression, classification, clustering, model evaluation
Deep learning Neural networks, CNNs, TensorFlow, Keras
Computer vision OpenCV preprocessing, image classification
Communication Case-study writeups, model interpretation, result summaries

Connection To Bioinformatics

The same modeling discipline used here applies directly to omics and translational research:

  • Define the biological or clinical question before modeling.
  • Separate signal from leakage, confounding, and batch effects.
  • Choose metrics that match the scientific risk.
  • Treat model output as evidence, not truth.
  • Make workflows reproducible enough for another scientist to inspect.

Credit

Program: MIT Professional Education - Applied Data Science: Leveraging AI for Effective Decision-Making.

This repository contains personal coursework notes and project artifacts. MIT and course instructors deserve credit for the program structure and teaching material; interpretations and portfolio framing here are my own.