This project contains a comprehensive analysis of Bitcoin price prediction using machine learning models and a comparative study between Bitcoin and Apple Inc. stock (AAPL) to understand the distinct characteristics of cryptocurrency and traditional stock markets.
This project aims to:
1-Predict Bitcoin prices using three machine learning models:
.Linear Regression: Provides a baseline prediction but lacks flexibility in capturing Bitcoin's volatility.
.Random Forest Regressor: Captures non-linear patterns with improved accuracy.
.LSTM (Long Short-Term Memory) Neural Network: Models sequential dependencies in time-series data for higher predictive accuracy.
2-Compare Bitcoin and Apple (AAPL) markets to illustrate their unique characteristics in terms of volatility, trading volume, and investor behavior in Tableau(over the last decade).
Data Sources:
Bitcoin Data:
.Historical Data (2014-2022): Collected from Kaggle’s Bitcoin historical data.
.Recent Data (2022-present): Sourced from the CryptoCompare API.
Apple Stock Data:
.Collected from Investing.com, covering the same time range as the Bitcoin dataset to enable a direct comparison.(https://www.investing.com/equities/apple-computer-inc-historical-data)
Model Selection and Insights:
.Linear Regression serves as a simple baseline model, but it struggles to capture Bitcoin’s non-linear and volatile price movements.
.Random Forest Regressor improves accuracy by capturing complex patterns.
.LSTM Model achieves the lowest RMSE, effectively modeling Bitcoin’s time-dependent trends, making it the preferred choice for price prediction.
Tableau Dashboard Overview:
1-Average Price: Bitcoin started with an average price of $344 in 2014, compared to Apple’s $28, with Bitcoin showing significantly more price volatility over time.
2-Standard Deviation: Bitcoin’s high standard deviation range (24 to 10,200) highlights its volatility, while Apple’s narrower range (0.6 to 21) illustrates a more stable growth pattern.
3-Correlation Between Price and Volume: Positive correlation in Bitcoin’s data suggests increased market interest with price rises, whereas Apple’s negative correlation reflects a stable, long-term investment appeal.
4-Total Volume Traded: Apple’s total traded volume, reaching 3.6 trillion dollars, showcases a deeper, more liquid market compared to Bitcoin’s volume of under 0.5 trillion dollars.
Conclusion:
The comparative analysis and prediction models reveal the stark contrasts between Bitcoin and Apple’s market behaviors. Apple’s stable growth and deep market make it attractive to traditional, risk-averse investors, while Bitcoin’s high volatility and rapid growth appeal to those willing to engage in a high-risk, high-reward environment. This project underscores the need for advanced modeling techniques, like LSTMs, to predict prices in volatile markets effectively. The LSTM model outperformed other methods, highlighting the need for sequential data modeling in volatile financial markets.
View the presentation here: https://docs.google.com/presentation/d/1c4GekwMa60iV31noQYOEVTvo2hz9lBrOiflla8sDcfU/edit#slide=id.p
Tableau public : https://public.tableau.com/app/profile/amir.h7728/viz/final22_17309089406480/Dashboard1?publish=yes