Stock Price Analysis Using EDA and Predictive Machine Learning Models

Project Overview

This project aims to predict stock prices by combining historical market data with advanced data processing techniques. We collect data for a diverse set of assets, including stocks, ETFs, commodities, and cryptocurrencies, and enrich this data with additional features. The ultimate goal is to create a robust dataset that can be used for quantitative analysis, trading strategies, and financial modeling.

Project Structure

StockPricePrediction/

├── data/ # Directory to store raw and enhanced data files ├── notebooks/ # Jupyter notebooks for data collection and analysis ├── scripts/ # Python scripts for data processing ├── .gitignore # Git ignore file ├── README.md # Project documentation ├── requirements.txt # List of Python libraries required for the project

Data Collection and Feature Engineering

Assets

We collect data for the following diverse assets:

• AAPL: Apple Inc.
• MSFT: Microsoft Corporation
• GOOGL: Alphabet Inc. (Google)
• AMZN: Amazon.com Inc.
• TSLA: Tesla Inc.
• SPY: SPDR S&P 500 ETF Trust (an ETF that tracks the S&P 500)
• GLD: SPDR Gold Shares (an ETF that tracks the price of gold)
• BTC-USD: Bitcoin in USD (Cryptocurrency)

Time Period

The data is collected for the period from January 1, 2010, to January 1, 2023.

Data Processing

Sure, here are the equations rewritten using more mathematical symbols:

1. Daily Returns: $R_t = \frac{P_t - P_{t-1}}{P_{t-1}}$

   where $R_t$ is the daily return, $P_t$ is the adjusted closing price at time $t$, and $P_{t-1}$ is the adjusted closing price at time $t-1$.

2. Moving Averages:

   • 20-day Moving Average (MA20): $MA_{20} = \frac{1}{20} \sum_{i=0}^{19} P_{t-i}$

   • 50-day Moving Average (MA50): $MA_{50} = \frac{1}{50} \sum_{i=0}^{49} P_{t-i}$

3. Volatility:

   • 20-day Volatility (standard deviation of daily returns): $\sigma_{20} = \sqrt{\frac{1}{20} \sum_{i=0}^{19} (R_{t-i} - \mu)^2}$

     where $\sigma_{20}$ is the 20-day volatility, $R_{t-i}$ is the daily return at time $t-i$, and $\mu$ is the mean daily return over the past 20 days.

Data Collection Script

The main script for collecting and processing data is provided in the notebooks directory under DataCollection.ipynb. This script performs the following steps:

Downloads historical data for each asset.
Adds extra features.
Saves both raw and enhanced data to CSV files.