This repository offers a curated collection of data science projects, primarily Jupyter notebooks and R analyses, designed as a reference for individuals learning data science fundamentals or building their own portfolios. It provides runnable examples across various domains, updated to function on a current Python 3.14 stack, enabling users to explore and adapt foundational data science techniques.

How It Works

The portfolio showcases projects spanning Machine Learning (supervised, unsupervised, reinforcement, deep learning), Natural Language Processing, and Data Analysis/Visualization. Core approaches involve applying standard algorithms like decision trees, CNNs (PyTorch), Q-Learning, PCA, Gaussian mixtures, XGBoost, and transformer models. Projects often include data preprocessing (ETL pipelines) and model evaluation, with some featuring web app integrations (Flask) or custom-built scoring mechanisms. The use of established libraries like Pandas, Scikit-learn, NLTK, and PyTorch ensures practical applicability.

Quick Start & Requirements

• Primary install/run: Use uv for environment management (uv sync then uv run jupyter lab) or standard pip (pip install -r requirements.txt).
• Prerequisites: Python 3.14. Some notebooks automatically download datasets (e.g., MNIST via torchvision, stock data via yfinance) on first run.
• Links: For current AI engineering work, see sajalsharma.com.

Highlighted Details

• Covers diverse ML tasks: predicting housing prices, donor identification, customer segmentation, autonomous driving agents, digit sequence recognition, disaster message classification, sentiment analysis, cross-language retrieval, and spam detection.
• Includes data analysis projects on walkability, stock market trends, election polls, and 911 calls, alongside R analyses of public health and social survey data.
• Features both traditional ML techniques (e.g., Naive Bayes, Logistic Regression, SVMs) and deep learning (PyTorch CNNs) and NLP approaches (NLTK, transformers).

Maintenance & Community

The repository has been refreshed for current Python versions. For current work and potential collaboration, contact contact@sajalsharma.com or visit sajalsharma.com. A "buy me a coffee" link is provided for support.

Licensing & Compatibility

No specific open-source license is mentioned in the provided README. Users should exercise caution regarding usage, modification, and distribution, especially for commercial purposes, until a license is clarified.

Limitations & Caveats

The projects originate from early career work (circa 2016), although they have been updated to run on modern Python stacks. The data included is explicitly stated as being for demonstration purposes only. A significant caveat is the absence of explicit licensing information, which may impact adoption decisions.