Upwork Project Data Analysis and Prediction Models

Overview

This project involves a comprehensive analysis of Upwork project data using Power BI to uncover actionable trends and insights, with a focus on predicting key project attributes such as budget, required skills, and completion timelines based on historical data.

Objectives

• Data Analysis: Analyzed Upwork project data using Power BI to identify trends and generate actionable insights.
• Prediction Models: Developed and compared four different machine learning models to predict project budgets, required skills, and completion timelines:
  • Recurrent Neural Networks (RNN)
  • Artificial Neural Networks (ANN)
  • Random Forest
  • Decision Tree
• Optimization: Improved model accuracy through advanced optimization techniques, particularly focusing on the RNN model.

Key Results

• Best Performing Model: RNN was identified as the top-performing model with an initial accuracy of 76%, which was optimized to 87% through advanced optimization techniques.
• Neural Networks: Utilized neural networks to model complex relationships in the project data, significantly outperforming traditional machine learning models like Decision Tree and Random Forest.

Technologies Used

• Power BI for data visualization and analysis
• Python for model development and optimization
• Libraries: TensorFlow, Keras, scikit-learn, pandas

Model Evaluation

• RNN: Best suited for sequential data, capturing the dependencies and trends in project data.
• ANN: Performed well for general-purpose prediction tasks.
• Random Forest: Useful for interpreting feature importance but less effective for time-based predictions.
• Decision Tree: Good for simple problems but lacked precision in this case.

Conclusion

Leveraging neural networks for complex forecasting tasks like project budgeting and timeline predictions has proven effective, providing improved accuracy compared to traditional machine learning methods.

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/upwork-project-analysis.git

2. Install necessary Python dependencies:

   pip install -r requirements.txt

3. Open the Power BI dashboard for interactive data analysis.
4. Run the Python scripts to train the models and optimize their performance.

Future Improvements

• Incorporate additional project features (e.g., client ratings, freelancer expertise) to further improve model accuracy.
• Experiment with more advanced deep learning architectures like LSTMs or Transformer models for better time-series prediction.