A comprehensive collection of AI/ML projects showcasing expertise in deep learning, NLP, reinforcement learning, graph-based algorithms, and data analysis from my Master's degree in AI.
This repository contains 10+ industry-focused projects spanning multiple AI/ML domains:
- Deep Learning: CNN-based image classification and time series forecasting
- Natural Language Processing: Transformer models, RAG architectures, and text classification
- Reinforcement Learning: Q-Learning and Monte Carlo implementations with interactive visualization
- Recommender Systems: Hybrid approaches combining graph neural networks with LLMs
- Graph Algorithms: Centrality measures, fraud detection networks, and pattern recognition
- Machine Learning: Classification, regression, and predictive modeling
Core Technologies: Python 3.8+, PyTorch, TensorFlow, Scikit-learn, Pandas, NumPy
Deep Learning: CNNs, Transformers, LSTM, RNNs, Attention mechanisms
NLP: LLMs, RAG pipelines, Transformers (HuggingFace), Vector embeddings
RL: Gymnasium, Q-Learning, Monte Carlo, Bellman equations
Recommender Systems: Graph Attention Networks (GAT), Collaborative filtering, Hybrid architectures
Graph Algorithms: NetworkX, Centrality measures (Betweenness, Closeness, Eigenvector, PageRank)
Visualization & APIs: Plotly, Matplotlib, FastAPI, Streamlit
Data Processing: Web scraping (BeautifulSoup), CSV analysis, SQL
Crop Disease Detection
CA01_CropDiseaseDetection_DL
CNN-based image classification for identifying crop diseases from plant leaf images. Demonstrates data preprocessing, model training, and evaluation on agricultural datasets.
Tech: PyTorch/TensorFlow, CNNs, Image classification | Key Learning: Transfer learning, data augmentation, class imbalance handling
Online Stock Forecasting
CA02_OnlineStockForecasting
Time series forecasting using LSTM networks for stock price prediction. Compares different temporal architectures and hyperparameter optimization.
Tech: LSTM, RNNs, Time series analysis, Pandas | Key Learning: Sequence modeling, feature engineering for temporal data
Blog Gender Classification & Transformer Paper Analysis
CA01_BlogGenderClassification_And_TransformerPaperAnalysis
Classification of blog authors by gender and comprehensive analysis of foundational transformer architecture papers.
Tech: NLP, Text classification, BERT, Transformers | Key Learning: Pre-trained models, NLP preprocessing, academic paper analysis
University RAG Chatbot
CA02_UniversityRAGChatbot
Full-stack RAG system for querying university documents using semantic search and LLM generation. Includes document ingestion pipeline and evaluation metrics.
Tech: LLMs, Vector embeddings, LangChain/RAG framework, Python | Key Learning: Information retrieval, prompt engineering, retrieval evaluation
Client Subscription Classifier
ClientSubscriptionClassifier
Binary classification of banking clients for subscription prediction. Includes feature engineering and model comparison techniques.
Tech: Scikit-learn, Random Forest, Logistic Regression | Key Learning: Class imbalance, feature selection, cross-validation
Energy Consumption Predictor
EnergyConsumptionPredictor
Regression model predicting household power consumption from temporal and statistical features on large-scale dataset (48MB+ data).
Tech: Regression, Time series, XGBoost, scikit-learn | Key Learning: Big data handling, time-based feature engineering, model optimization
Taxi Grid Navigation System (Prototype)
TaxiGridNavigationSystem_Prototype
Full-stack interactive RL teaching tool with FastAPI backend and React frontend. Implements Q-Learning and Monte Carlo algorithms with real-time visualization.
Tech: FastAPI, Gymnasium, React 18, TypeScript, Vite | Key Learning: RL algorithms, full-stack development, state machine design
Taxi Grid Navigation System (Using RL)
TaxiGridNavigationSystem_Using_RL
Implementation of Q-Learning and Monte Carlo algorithms for optimal taxi routing in grid environments.
Tech: Q-Learning, Monte Carlo, Gymnasium, Python | Key Learning: Value function approximation, exploration-exploitation trade-off
GAT + LLM Based Movie Recommender System
GAT_LLM_Based_MovieRecommenderSystem
Hybrid recommendation system combining Graph Attention Networks for user-item relationships with LLMs for personalized content generation.
Tech: Graph Neural Networks (GAT), LLMs, PyTorch Geometric | Key Learning: Hybrid recommenders, graph-based collaborative filtering
LinkedIn Jobs Recommender System
LinkedinJobs_RecommenderSystems
Job recommendation engine using similarity metrics and ranking algorithms on scraped LinkedIn data.
Tech: Web scraping, Collaborative filtering, NLP | Key Learning: Recommendation algorithms, data pipeline design
Centrality Measures Calculation
CentralityMeasuresCalculationTasks
Analysis of network centrality metrics (Degree, Betweenness, Closeness, Eigenvector, PageRank) with Neo4j graph database and comprehensive visualizations.
Tech: Neo4j, NetworkX, Graph theory, Cypher | Key Learning: Graph algorithms, network analysis
Fraud Detection with Graph Analysis
Fraud Detection
Network-based fraud detection using graph algorithms to identify suspicious transaction patterns and relationships.
Tech: Graph analysis, NetworkX, Neo4j, Cypher | Key Learning: Anomaly detection, network patterns
LinkedIn Jobs Scraper & Data Analysis
LinkedinJobsScraper_DataAnalysis
Web scraping, exploratory data analysis, and insights generation from LinkedIn job postings including trend analysis and market intelligence.
Tech: BeautifulSoup, Pandas, Data visualization | Key Learning: Web scraping, EDA, business intelligence
- Python 3.8 or higher
- pip or conda package manager
- Git
# Clone the repository
git clone <repo_url>
cd MS_In_AI_Projects
# Create a virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies (varies by project)
pip install -r requirements.txt # if available in project subdirectoryJupyter Notebooks (Deep Learning, NLP, ML, Recommender Systems):
cd <project_directory>
jupyter notebook <notebook_name>.ipynbPython Scripts (RL, specific implementations):
cd <project_directory>
python <script_name>.pyFull-Stack Applications (Taxi RL Simulation, RAG Chatbot):
- See project-specific README or documentation for detailed setup instructions
- Each project may have its own dependencies—check subdirectory for
requirements.txt - Some projects require API keys (LLM services) or large datasets
- Run from project subdirectory for relative path configurations
Name: Bhargava Koya
Email: [bhargavakoya77@example.com]
LinkedIn: https://www.linkedin.com/in/bhargava-koya7/
GitHub: https://github.com/bhargavkoya
Master's in AI