ML Algorithm Tutor with Langchain

A comprehensive, interactive Machine Learning tutoring chatbot built with Streamlit and Google's Gemini AI, featuring a modular OOP architecture, hands-on practice problems, code execution, and rich visualizations.

Architecture Overview

The application is organized into modular, object-oriented components:

/app Package Structure:

• ui.py - Main UI class handling all Streamlit components and user interactions
• curriculum.py - Curriculum management with ML topics, categories, learning paths, and progress tracking
• tutor.py - AI tutor logic using Google's Gemini model for explanations, practice problems, and code review
• memory.py - Memory management for chat history and session state
• practice_problems.py - Generates quizzes, coding, and dataset problems for ML topics
• code_executor.py - Securely executes and evaluates user-submitted code
• visualizations.py - Interactive visualizations for ML algorithms using Plotly and Matplotlib

Entry Points:

• main.py - Main application entry point (runs the Streamlit app)
• run_app.py - Enhanced runner with dependency checking and setup

Features

• Interactive ML Curriculum: Browse topics by category, difficulty, and learning path
• AI-Powered Explanations: Get detailed, structured explanations with code examples
• Practice Problems: Solve quizzes, coding challenges, and dataset-based problems for hands-on learning
• Code Execution & Feedback: Write, run, and get instant feedback on your code
• Streaming Responses: Real-time AI response generation
• Visualizations: Interactive charts and algorithm visualizations for deeper understanding
• Progress Tracking: Track your learning progress, quiz scores, and completed topics
• Chat History: Review your learning conversation and clear history as needed
• Responsive UI: Clean, modern interface with custom styling

Installation & Setup

1. Clone and navigate to the project:

   git clone <your-repo-url>
   cd ML-LangBot

2. Install dependencies:

   pip install -r requirements.txt

3. Set up environment variables: Create a .env file in the project root with your Google API key:

   GOOGLE_API_KEY=your_actual_google_api_key_here

4. Run the application:

   python run_app.py

   Or directly:

   streamlit run main.py

The app will open in your browser at http://localhost:8501

Usage

1. Browse Curriculum: Use the sidebar to explore ML topics by category, difficulty, or learning path
2. Ask Questions: Type any ML algorithm or topic in the input field to get an AI-powered explanation
3. Practice Mode: Enable practice mode to solve quizzes, coding, and dataset problems for selected topics
4. Code Execution: Write and evaluate your code for coding challenges, with instant feedback
5. Visualizations: View interactive visualizations for supported algorithms
6. Track Progress: Monitor your progress, quiz scores, and completed topics
7. View/Reset History: Review or clear your chat history at any time

Supported ML Topics

• Data Preprocessing: Missing Data, Encoding, Feature Scaling
• Regression: Linear Regression, Polynomial Regression, SVR, Decision Tree Regression, Random Forest Regression
• Classification: Logistic Regression, K-NN, SVM, Naive Bayes, Decision Tree Classification, Random Forest Classification
• Clustering: K-Means, Hierarchical Clustering
• Association Rule Learning: Apriori, Eclat
• Reinforcement Learning: UCB, Thompson Sampling, Q-Learning
• NLP: Bag-of-Words, TF-IDF, Word2Vec, BERT
• Deep Learning: ANN, CNN, RNN, LSTM
• Dimensionality Reduction: PCA, LDA, Kernel PCA
• Model Selection & Boosting: Cross Validation, Grid Search, XGBoost, AdaBoost

Note: Practice problems and visualizations are available for a subset of topics. More are being added!

Contributing

The modular OOP structure makes it easy to contribute:

• Add new curriculum topics or learning paths in curriculum.py
• Enhance UI components in ui.py
• Improve AI explanations or practice problems in tutor.py and practice_problems.py
• Extend code execution or feedback in code_executor.py
• Add or improve visualizations in visualizations.py
• Report bugs or suggest features via issues or pull requests

Requirements

See requirements.txt for all dependencies. Key packages:

• streamlit, langchain, langchain-google-genai, google-generativeai, python-dotenv
• plotly, matplotlib, seaborn, numpy, pandas, scikit-learn, xgboost

Deployment

https://ml-langbot.streamlit.app