Predicting Global Stock Prices and Analyzing Stock-GDP Interplay

Overview:
This project explored the relationship between global stock market trends and GDP fluctuations over time. Using advanced data analytics, machine learning techniques, and predictive modeling, the project identified key patterns in the financial market and developed a robust forecasting framework for stock prices. The analysis provided actionable insights for investors, policymakers, and financial analysts.

​

Key Contributions:

1. Data Preprocessing and Integration:

   • Sourced stock market data from Yahoo Finance and GDP per capita data from global repositories such as the World Bank and IMF.

   • Handled missing values, normalized datasets, and added derived features like USD-converted stock closing prices for cross-comparative analysis.

2. Exploratory Data Analysis (EDA):

   • Conducted in-depth analysis of stock and GDP data to uncover trends and patterns.

   • Visualized stock price trends using line graphs, cumulative price comparisons, and volume bars.

   • Highlighted the economic impact of events like the 2020 COVID-19 crisis on global markets and GDP.

3. Feature Engineering and Selection:

   • Used Fast Fourier Transform (FFT) for time-series frequency analysis and feature extraction.

   • Applied Gradient Boosting and Random Forest algorithms for feature selection, identifying key drivers of stock price changes with precision.

4. Machine Learning Models:

   • Developed and evaluated multiple regression and clustering models, including Linear Regression, Gradient Boosting, K-Means, and Decision Tree.

   • Optimized Gradient Boosting, achieving the lowest Mean Squared Error (MSE) for accurate stock price predictions.

5. Time-Series Forecasting:

   • Forecasted stock prices using FFT-based time-series models, leveraging historical data to predict future trends.

   • Clustered financial data using K-Means to group stock indices for enhanced pattern recognition.

6. Comparative Analysis:

   • Correlated stock market trends with GDP fluctuations to highlight the interconnectedness of economic indicators.

   • Demonstrated the impact of stock market volatility on GDP performance and vice versa.

 

Results:

• Achieved precise stock price predictions with advanced machine learning models.

• Provided insights into the economic effects of stock market trends on global GDP fluctuations.

• Visualized key financial patterns using clear and impactful data visualizations.

 

Skills and Technologies:

• Programming & Tools: Python, Pandas, NumPy, Matplotlib, Seaborn

• Machine Learning: Gradient Boosting, Random Forest, Linear Regression, FFT, K-Means Clustering

• Data Processing: Feature engineering, handling missing values, data normalization, dimensionality reduction

• Visualization: Time-series plots, line graphs, bar charts, and cluster visualizations

 

Impact:
This project demonstrates the ability to integrate data analytics and machine learning for real-world financial applications. It highlights expertise in predictive modeling, data-driven decision-making, and creating actionable insights, aligning directly with the roles of data analysts, data scientists, and machine learning professionals.