Stocks Analysis and Forecasting Application

---

Overview

This application is a comprehensive stock analysis and forecasting tool. It allows users to visualize historical data, forecast future stock prices using multiple advanced models, and interact with data through a sleek graphical user interface (GUI).

Features

1. Real-Time Stock Price Fetching: Fetches the latest stock prices using Yahoo Finance.
2. Historical Data Visualization: Displays historical stock data for the last 10 years.
3. Forecasting Models:
   • ARIMA (Auto-Regressive Integrated Moving Average)
   • ETS (Exponential Smoothing)
   • Random Forest
   • LSTM (Long Short-Term Memory)
4. Interactive Graphs:
   • Displays model forecasts for the next 12 months.
   • Hover functionality to show predicted price and date.
5. Modern GUI:
   • User-friendly interface..
   • Interactive graphs with zoom and pan capabilities.
6. Multi-Model Comparison:
   • Compare forecasts from different models on the same graph.

---

Application Workflow

1. Data Fetching:
   • Fetches historical stock data from Yahoo Finance for analysis.
   • Updates real-time prices using intraday data.
2. Data Preprocessing:
   • Cleans and prepares data for models.
   • Handles missing data, converts formats, and ensures compatibility.
3. Forecast Generation:
   • Each model generates predictions for the next 12 months.
   • Predictions start from the current date.
4. Visualization:
   • Forecasts are displayed on a graph.
   • Users can view specific predictions by hovering over points on the graph.

---

Modules and Functions

1. Data Fetching

Function: fetch_stock_data(stock_symbol, start_date, end_date)

• Fetches historical stock data from Yahoo Finance.
• Returns a time-series dataset for analysis.

Function: fetch_intraday_price(stock_symbol)

• Fetches the latest stock price using intraday data.

---

2. Forecasting Models

• ARIMA (AutoRegressive Integrated Moving Average):
  • Provides statistical-based predictions.
  • Suitable for time-series data with trends.
• ETS (Exponential Smoothing):
  • Uses exponential smoothing for seasonal data.
• Random Forest:
  • A machine learning model that incorporates lag features and technical indicators.
• LSTM (Long Short-Term Memory):
  • A neural network-based model for sequential data.

---

3. Graphical User Interface (GUI)

Function:

• Launches the main GUI window.
• Displays historical data and model forecasts.
• Features:
  • Interactive graph with hover functionality.
  • Buttons to show historical data and exit.

Function:

• Opens a new window displaying historical data for the last 10 years.

---

Dependencies

Python Libraries

• yfinance: Fetches stock data.
• pandas: Data manipulation.
• matplotlib: Visualization.
• tensorflow: Neural network models (LSTM).
• statsmodels: ARIMA and ETS models.
• sklearn: Random Forest and preprocessing tools.
• mplcursors: Adds hover functionality.
• tkinter: GUI framework.

Installation

pip install yfinance pandas matplotlib tensorflow statsmodels scikit-learn mplcursors tk

---

How to Use

1. Launch the Application:

   • Run the script using python stocks_forecast.py.

2. Select Stock:

   • The application fetches data for IBM stocks by default.
   • Modify the stock_symbol variable to analyze other stocks.

3. View Forecasts:

   • Observe forecasts from ARIMA, ETS, Random Forest, and LSTM models.
   • Hover over points on the graph to view predictions.

4. Exit:

   • Click the "Exit" button to close the application.

---

Customization

Modify Stock Symbol

Change the stock_symbol variable in the __main__ section:

stock_symbol = "AAPL"  # For Apple stock

Adjust Forecast Horizon

Modify the steps parameter in model functions to increase/decrease the forecast period.

Add More Models

Additional models can be implemented e.g., XGBoost, Prophet.

---

Future Improvements

1. Additional Models: Add models like Prophet, XGBoost, and SARIMA.
2. User Preferences: Allow users to select models and forecast durations dynamically.
3. Data Export: Functionality to save forecasts as CSV or Excel files.

---

Known Issues

1. LSTM Predictions:
   • Requires sufficient data for accurate predictions.
   • Training may be slow for larger datasets.
2. ARIMA Limitations:
   • May not handle non-stationary data well without preprocessing.
3. Real-Time Price Delay:
   • Intraday data from Yahoo Finance may have a delay of up to 30 minutes.

---

---

Support

For issues or suggestions, contact: Developer: Atanas Youdanov
Email: [email protected] GitHub: GitHub Profile

---