Market Capitalization Forecasting

Overview

This project provides a comprehensive pipeline for forecasting market capitalization of companies using various time series forecasting techniques, including:

• Transformer-based models (TimeSeriesTransformer)
• ARIMA models
• RNN models

The system fetches market cap data from Yahoo Finance, processes it into a time series format, trains multiple forecasting models, and generates predictions with evaluation metrics.

Project Structure

marketcap-forecasting/
├── data_preparation/
│   ├── fetch_data.py            # Fetches market cap data from Yahoo Finance
│   ├── prepare_dataset.py       # Processes raw data into time series format
│   └── check_for_nan.py         # Validates data quality
├── forecasting/
│   ├── train.py                 # Trains the TimeSeriesTransformer model
│   ├── test_model.py            # Evaluates the trained model
│   ├── forecast_future.py       # Generates future forecasts
│   ├── ARIMA.py                 # ARIMA model implementation
│   └── RNN.py                   # RNN model implementation
├── visualization/
│   └── plot_marketcap_data.py   # Visualizes market cap trends
├── market_cap_data/             # Directory for raw market cap data
├── prepared_marketcap_dataset/  # Processed dataset directory
├── saved_models/                # Saved models directory
└── saved_forecasts/             # Saved forecasting results

Key Features

1. Data Pipeline

   • Automated data fetching from Yahoo Finance
   • Time series preprocessing and validation
   • Static feature encoding (sector, industry, market cap category)

2. Modeling Approaches

   • State-of-the-art Time Series Transformer
   • Traditional ARIMA models
   • Recurrent Neural Networks (RNNs)

3. Evaluation Metrics

   • MASE (Mean Absolute Scaled Error)
   • sMAPE (Symmetric Mean Absolute Percentage Error)
   • MAPE (Mean Absolute Percentage Error)

4. Visualization

   • Interactive forecast plots
   • Model performance comparisons
   • Time series trend analysis

Getting Started

Prerequisites

• Python 3.8+
• Required packages: pip install -r requirements.txt

Basic Usage

1. Fetch Data

   python fetch_data.py

2. Prepare Dataset

   python prepare_dataset.py

3. Train Model

   python train.py

4. Evaluate Model

   python test_model.py

5. Generate Forecasts

   python forecast_future.py

Model Comparison

Model Type	Strengths	Best For
TimeSeriesTransformer	Captures complex patterns, handles multiple series	Large-scale forecasting
ARIMA	Simple, interpretable	Short-term forecasts
RNN	Handles sequential dependencies	Medium-term forecasts

Results

Example forecast for Apple Inc.:

License

This project is licensed under the MIT License - see the LICENSE file for details.