Skip to content

Repository files navigation

🛢️ OilyGiant — Optimal Oil Well Location Prediction

📘 Overview

This project helps OilyGiant Mining Company identify the most profitable region for developing new oil wells.
Using geological data from three regions, a Linear Regression model predicts oil reserves, estimates potential profits, and evaluates investment risks using bootstrapping.


🚀 Project Steps

1. Data Preparation

  • Load and preprocess geological data for three regions:
    • geo_data_0.csv
    • geo_data_1.csv
    • geo_data_2.csv
  • Each dataset contains:
    • id — unique oil well identifier
    • f0, f1, f2 — geological features
    • product — actual oil reserve volume (thousand barrels)

2. Model Training

For each region:

  • Split the data into 75% training and 25% validation sets.
  • Train a Linear Regression model to predict oil reserves.
  • Evaluate model performance using RMSE and the average predicted reserve volume.

3. Profit Calculation

  • Budget: $100 million for 200 wells.
  • Revenue per thousand barrels: $4,500.
  • Select the top 200 wells with the highest predicted reserves.
  • Calculate the total profit for each region.

4. Risk Assessment

  • Apply Bootstrapping (1,000 samples) to estimate:
    • Average profit
    • 95% confidence interval
    • Probability of loss (negative profit)
  • Only regions with a risk of loss < 2.5% are considered.
  • Choose the region with the highest mean profit and acceptable risk.

🛠️ Tools & Libraries

  • Python
  • Pandas
  • NumPy
  • Scikit-learn
  • Matplotlib
  • Seaborn

📈 Outcome

The project delivers a data-driven recommendation for selecting the optimal region to develop new oil wells — maximizing profitability while minimizing financial risk.


📁 Data

File Description
geo_data_0.csv Geological data for region 0
geo_data_1.csv Geological data for region 1
geo_data_2.csv Geological data for region 2

⚠️ All datasets are synthetic and provided for educational and analytical purposes.


🧮 Key Assumptions

  • Only Linear Regression is used for predictions.
  • Each exploration studies 500 points, and the best 200 are developed.
  • Development budget: $100 million.
  • Revenue per unit (thousand barrels): $4,500.

📊 Results

  • Predicted oil reserves and RMSE per region.
  • Estimated profits with confidence intervals.
  • Final recommendation for the most profitable, lowest-risk region.

🏁 Conclusion

This analysis enables OilyGiant to make a data-informed investment decision, identifying the region with the highest expected profit and minimal financial risk.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages