📈 Real-Time Stock Price Prediction and Market Analysis using Machine Learning

This project is a real-time stock market prediction and analysis web app built using Python, Keras, and Streamlit. It predicts future stock prices using a deep learning model trained on historical stock data fetched from Stooq.

---

🎥 Demo

📽️ Click here to watch the demo video

---

🧩 Problem Statement

• Stock price movements are highly volatile and influenced by complex patterns, making accurate forecasting challenging for investors and analysts.
• Traditional forecasting methods often fail to capture the non-linear trends and time-dependent nature of stock market data.
• There is a need for an intelligent, real-time system that can predict future stock prices using deep learning techniques, based on historical trends and patterns.
• This project aims to build a web-based application that leverages an LSTM model to forecast stock prices and visualize trends, helping users make more informed decisions.

---

🧾 Existing System

• Uses statistical models, technical indicators, and fundamental analysis.
• Limited in handling complex patterns and time-based dependencies in stock data.
• Not well-suited for real-time market analysis.
• Struggles to adapt to fast-changing market conditions

---

💡 Proposed System

• A machine learning-based approach using an LSTM model trained on historical stock prices.
• Utilizes free data from Stooq for accessibility.
• Predicts future prices and visualizes them with moving averages and comparison plots.
• Fully interactive and user-friendly interface built with Streamlit.

---

Methodology

1. Data Collection
2. Data Preprocessing
3. Feature Engineering
4. Model Training
5. Stock Price Prediction
6. Visualization and Reporting

---

🔍 Key Features

• 📊 Fetches historical stock data from 2010–2023 for any ticker (default: AAPL) using pandas_datareader.
• 📉 Visualizes stock price trends with moving averages (100MA & 200MA).
• 🤖 Predicts future stock prices using a pre-trained LSTM model (keras_model.h5).
• 📅 Displays predicted prices for the next 10 days using recent trends.
• 📈 Compares actual vs. predicted prices using interactive line plots.
• 🌐 Fully interactive web interface built using Streamlit.

---

🧠 Technologies & Libraries Used

• Python
• TensorFlow / Keras (LSTM model)
• Streamlit (for real-time UI)
• pandas (for data handling)
• matplotlib (for visualizations)
• scikit-learn (MinMaxScaler for normalization)
• pandas_datareader (fetching stock data from Stooq)

---

🧪 How It Works

1. User inputs a stock ticker (e.g., AAPL).
2. App fetches historical data (2010–2023) from Stooq.
3. Visualizes:
   • Raw closing prices
   • 100-day & 200-day moving averages
4. Splits data into:
   • 70% training
   • 30% testing
5. Uses an LSTM model to:
   • Predict prices
   • Compare with actual values
   • Forecast the next 10 days
6. All results are visualized and displayed on a real-time Streamlit dashboard.

---

🎯 Model Accuracy

While no explicit metric (like RMSE or R²) is shown in the app, the LSTM model demonstrates approximately 85% trend accuracy, based on visual comparison of actual vs. predicted curves.

---

📌 How to Run

1. Clone the repository: Clone the project repository to your local machine using Git (or download as ZIP): git clone <repo_url>

2. Set up a virtual environment: Open a terminal and navigate to the project folder. Then create and activate a virtual environment:

For Windows: python -m venv venv venv\Scripts\activate

For macOS/Linux: python3 -m venv venv source venv/bin/activate

3. Install required dependencies: Once your virtual environment is active, install the required packages using the requirements.txt file: pip install -r requirements.txt

4. Run the Streamlit app: To start the Streamlit web app, run the following command in your terminal: streamlit run app.py

5. Access the web app: After running the command, Streamlit will automatically open the app in your default web browser. If not, you can manually open the following URL in your browser: http://localhost:8501

6. Enter a Stock Ticker: In the web interface, you can enter any stock ticker (e.g., AAPL) into the search bar. The app will fetch and display: Historical stock data (from 2010 to 2023). Stock price predictions for the next 10 days.