aaryansamanta/advanced-ai-curriculum

Portfolio of capstone projects from AI & Technology, Data Science, AI Internship, Machine Learning and AI Humanities Honors. CNN pneumonia detection on chest X-rays (>90% accuracy), healthcare EDA, diabetes risk prediction and AI ethics research. Python, scikit-learn, TensorFlow/Keras.

What it solves

This repository solves the problem of presenting original, unpublished capstone work from multiple advanced AI and Data Science programs in one place. It lets the author demonstrate hands-on skills in machine learning, deep learning, NLP, and healthcare analytics without exposing sensitive datasets or compromising academic integrity.

How it works

The repository is organized into five course folders—AI Internship Honors, Data Science Honors, AI Technology Honors, AI Humanities Honors, and Machine Learning Honors—each containing numbered assignments and a capstone project. Every project has its own README explaining the dataset, methodology, and takeaways, along with source code in Jupyter notebooks or Python files. The featured capstones cover pneumonia detection with CNNs, disease symptom analysis, diabetes prediction, and an AI ethics report.

Who it's for

This is built for recruiters, educators, and fellow learners who want to see applied AI work in healthcare and data science. It also serves as a reference for how to structure portfolio-capstone work while keeping unpublished research protected.

Highlights

  • CNN-based pneumonia detection from chest X-rays with a full technical report and slide deck
  • End-to-end data science capstone analyzing a real healthcare dataset for disease symptoms and patient profiles
  • Complete machine learning pipeline for diabetes risk prediction, including a walkthrough video
  • AI ethics and society capstone examining the societal impact of AI, delivered as a report and presentation

Related

  • Project
    ncoevoet/facet

    Facet is a private, self-hosted photo manager that scores each image across 9 quality dimensions, then offers a web gallery with culling, face recognition, semantic search, sharing links, and optional GPU-accelerated AI — all offline with no cloud accounts.

  • Project
    myBoris/wzry_ai

    An open-source AI model project designed to automate gameplay for Honor of Kings using machine learning and custom screen coordinate mapping.

  • Project
    yakhyo/face-parsing

    A computer vision tool that uses a BiSeNet architecture to segment facial features like eyes, nose, and skin, enabling applications in digital makeup, AR, and emotion detection.

  • Project
    Sompote/DINOV3-YOLOV12

    A hybrid object detection framework combining YOLOv12 and DINOv3 to provide superior accuracy and faster convergence, especially on small or complex datasets.