StockSight – Stock Trend Prediction

A Flask-based web application for downloading stock data, performing exploratory data analysis (EDA), visualizing price trends (including candlestick charts and moving averages), and predicting future stock prices using LSTM (Long Short-Term Memory) neural networks.

🎬 Demo

🔍 Overview

This project demonstrates a full end-to-end pipeline for stock trend prediction:

1. Data Acquisition: Download historical price data from Yahoo Finance using yfinance.
2. Exploratory Data Analysis: Display raw data, descriptive statistics, and visualize:
   • Line charts of Open, High, Low, Close, Volume
   • Candlestick plots
   • Simple Moving Averages (SMA) and Exponential Moving Averages (EMA)
3. Modeling: Prepare data with MinMaxScaler, split into training/testing sets, then train an LSTM model to forecast future prices.
4. Web Interface: Use Flask to build a front-end where users enter a stock ticker to:
   • View EDA charts
   • Download the raw CSV
   • See predictions overlaid with actual prices

💡 Key Features

• Download & Preview Data: Fetch any public stock by ticker symbol.
• Descriptive Statistics: Summary statistics (mean, std, percentiles).
• Interactive Visualizations:
  • Candlestick charts with Plotly
  • Line plots for OHLC & Volume
  • SMA & EMA overlays
• LSTM-based Forecasting: Predict next-day or multi-day prices.
• Model Persistence: Save/load trained model (.h5 file) to avoid retraining.
• CSV Export: Download the processed dataset.

🛠️ Tech Stack

• Language: Python 3.x
• Web Framework: Flask
• Data Handling: Pandas, NumPy
• Visualization: Matplotlib, Plotly
• Finance API: yfinance
• Machine Learning:
  • Preprocessing: Scikit-learn (MinMaxScaler)
  • Deep Learning: TensorFlow / Keras (Sequential API, LSTM layers)
• Template Rendering: Jinja2 (Flask)

🚀 Installation & Setup

1. Clone the Repository

   git clone https://github.com/namaniisc/StockSight.git
   cd StockSight

2. Create & Activate a Virtual Environment

   python3 -m venv venv
   source venv/bin/activate       # macOS/Linux
   venv\Scripts\activate        # Windows

3. Install Dependencies

   pip install -r requirements.txt

4. Download Pretrained Model (optional)

   If you wish to skip training, download stock_dl_model.h5 into the project root.

📂 Project Structure

├── app.py                   # Flask application
├── models/                  # Saved Keras model (.h5)
├── static/                  # Generated plots & downloaded CSVs
├── templates/
│   └── index.html           # Main UI template
├── requirements.txt         # Python dependencies
├── README.md                # Project documentation
└── notebooks/               # Jupyter notebooks used for prototyping

⚙️ Usage

1. Run the Flask App

   flask run      # or python app.py

2. Interact

   • Open your browser to http://127.0.0.1:5000
   • Enter a stock ticker (e.g., AAPL, POWERGRID.NS)
   • View EDA charts, download CSV, and see LSTM forecasts.

📊 Model Training (Optional)

If you want to retrain the LSTM model:

1. Open notebooks/train_model.ipynb.
2. Adjust hyperparameters (e.g., lookback window, epochs).
3. Run all cells to preprocess data, train the model, and save stock_dl_model.h5.

🤝 Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository.
2. Create a new branch: git checkout -b feature/YourFeature.
3. Commit your changes: git commit -m 'Add some feature'.
4. Push to the branch: git push origin feature/YourFeature.
5. Open a Pull Request.