Nigerian Stock Market Price Prediction

This project predicts stock prices for companies on the Nigerian Stock Exchange using machine learning. Users can select a stock ticker and see price predictions for future days through an easy-to-use web app.

use

Features

• 📈 Predict future stock prices based on historical data
• 🏢 Support for multiple Nigerian companies
• 📊 Interactive charts showing historical prices and predictions
• 📅 Forecasts starting from today for any number of days ahead
• 💹 Detailed prediction tables with dates and prices

How It Works

1. Data Collection and Cleaning

We collect Nigerian stock market data with these fields:

• Date
• Open price
• High price
• Low price
• Close price
• Volume
• Percentage change

The data is cleaned by:

• Removing unnecessary columns
• Converting text to proper numbers
• Handling missing values
• Converting volume notations (M for million, K for thousand)

2. Training Models

For each company (ticker):

1. We split the data into training (80%) and testing (20%) sets
2. We train a Linear Regression model using these features:
   • Open price
   • High price
   • Low price
   • Trading volume
   • Percentage change
3. The model learns to predict the closing price
4. We save each trained model as a file

3. Web Application

Our Streamlit app allows users to:

1. Select a company from a dropdown menu
2. Choose how many days to predict (1-30)
3. Click "Predict" to see future price estimates
4. View results as both a chart and a table

How to Run the Project

Requirements

• Python 3.7 or higher
• Required packages listed in requirements.txt

Setup

1. Clone this repository

git clone <repository-url>

2. Install required packages

pip install -r requirements.txt

3. Run the Streamlit app

streamlit run app.py

Project Structure

Nigerian Stock Market Price Prediction/
├── data/
│   └── Nigerian_stock_market.csv    # Stock market data
├── models/                          # Trained models
│   ├── AIICO_model.pkl
│   ├── DANGCEM_model.pkl
│   └── ...
├── app.py                           # Streamlit web application
|___ Notebooks                       #Notebook Folder
└── requirements.txt                 # Project dependencies

Usage Example

1. Open the app in your web browser
2. Select a stock ticker (e.g., "DANGCEM")
3. Use the slider to select how many days to forecast (e.g., 7 days)
4. Click "Predict Prices" button
5. View the historical data and predictions chart
6. Check the detailed forecast table below the chart

Future Improvements

• Add more advanced prediction models
• Include more Nigerian stocks
• Add confidence intervals to predictions
• Provide trading signals (buy/sell/hold recommendations)
• Automatic data updates

Credits

• Stock data sourced from Nigerian Stock Exchange
• Built with Python, Streamlit, and scikit-learn