🚀 Data Science Practice – Cryptocurrency Analysis

🔍 First Hands-On Practice in Data Science

This project explores financial data to analyze relationships between popular cryptocurrencies. It covers data collection, processing, and advanced statistical techniques to uncover insights into market behavior.

---

📌 Key Topics

📊 1. Data Collection, Processing & Resampling

• 🔹 Data Collection: Fetching OHLCV Candlestick Data
  • Modular Python functions for API interactions
  • Processing OHLCV data into a representative price series
  • Extracting implied USDT-TMN price series
• 🔹 Resampling:
  • Selection of time scales
  • Methodological approach
• 🔹 Handling Market Anomalies:
  • Missing data management
  • Outlier detection and correction
  • Data integrity assurance

📈 2. Exploratory Data Analysis (EDA)

• 📌 Log Returns, Volatility & Normality Assessment:
  • Volatility estimation & clustering (EWMA)
  • Statistical summaries
  • Graphical & quantitative normality tests
  • Importance of normality in financial models
• 📌 Autocorrelation & Stationarity Analysis:
  • ACF & PACF plots
  • Stationarity testing
  • Non-stationarity & autocorrelation interplay
• 📌 Inter-Market Analysis:
  • Synchronous & lagged correlations
  • Strategic application

🔗 3. Cointegration Analysis

• ✅ Cointegration testing methodology
• ✅ Dynamic analysis of cointegration parameters

📉 4. Error Correction Model (ECM)

• 🔄 ECM development
• 📊 Analysis of reversion dynamics

---

📌 Why This Matters?
Understanding market trends and price relationships is crucial for developing trading strategies and risk management in the crypto space. This project provides a structured approach to analyzing cryptocurrency data using statistical and econometric methods.

---

⚠️ Note: This project is my first experience in data science, and I acknowledge that it may have various shortcomings. I warmly welcome any collaboration, feedback, and suggestions to improve it. Your insights would be greatly appreciated! Also, if you need datasets, you can contact me