Predict Purchase Behavior From Market Data

Job ID: 39716810

Budget: $8 – $15 USD

I will hand over a structured set of historical market data and need you to turn it into a working model that reliably forecasts customer purchase behavior. The task starts with cleaning and exploring the data, moves through feature engineering, then lands on building and validating a predictive algorithm.

Use whatever stack you are most comfortable with—Python (pandas, scikit-learn, XGBoost) or R (tidyverse, caret) are both fine—as long as the workflow is reproducible and well-documented. Once the model is tuned, I also want a concise write-up that explains the key drivers you uncovered, the evaluation metrics you chose, and a short guide on how I can run or retrain the model in the future.

Deliverables
• Cleaned and labeled dataset
• Fully commented source code / notebook
• Trained model file plus environment specs
• Summary report with insight visuals and predictive accuracy figures

I’m aiming for a solution I can plug into future market data drops, so please keep modularity and clarity in mind.