Corporate Entity Tracking Automation>

Updated: April 2, 2024

Machine learning research project at the University of Central Florida (UCF)

CETA focuses on analyzing and understanding global corporate entities through a large historical dataset with records dating as far back as 1930. Objectives include tracking foreign ownership, influence, and illegal activities, while making temporal predictions regarding the longevity of companies based on their attributes.

Project Overview

The cornerstone of my contribution is the development of a custom RandomForestRegressor model, designed to predict how long a company is likely to last based on its attributes. I'm continually fine-tuning the model for improved accuracy and expanded capabilities.

Tools and Technologies Used

RandomForestRegressor: Machine learning model for predictive analysis.
NetworkX: Creating and analyzing corporate entity graphs and relationships.
Pandas: Data manipulation and analysis.
scikit-learn: Building and optimizing the RandomForestRegressor model.
GraphML: Handling graph data in a structured format.
NumPy: Efficient numerical operations.

Key Research Focus

Temporal Predictions: Predicting corporate longevity using ownership, financials, and history.
Foreign Ownership and Influence: Assessing how foreign entities affect corporate structures.
Illegal Activities: Identifying patterns indicating potential unlawful operations.

Model Development

I created the RandomForestRegressor model from scratch and continuously refine it. Improvements include feature engineering, hyperparameter tuning, and validation with various machine learning techniques to ensure higher predictive accuracy and broader utility.

Getting Started

Clone the repository.
Install the required Python libraries (Pandas, NetworkX, scikit-learn, NumPy, etc.).
Load the dataset and GraphML files into your environment.
Run the provided scripts to execute the model and analyze the results.

Contributing

This project is a work in progress, and contributions are welcome. Whether improving the model, adding new features, or exploring new angles of the data, your collaboration can drive its advancement.