Geopolitics-oil-price-model

Predicting next-month crude oil price movements using historical data and geopolitical conflict signals between israel and iran.

• Data: Monthly oil prices from 1983 to 2025 from https://www.kaggle.com/datasets/sc231997/crude-oil-price/data
• Features:
  • Historical price
  • Monthly price change
  • Conflict indicator (binary feature: 1 if known conflict occurred during the month)
• Target:
  • 0: Price drops >3%
  • 1: Price remains within ±3%
  • 2: Price rises >3%

The model predicts the next month's price trend, which is visualized alongside historical data.

📈 Model

We use a RandomForestClassifier to classify the next-month price trend into three categories. The model is trained and evaluated using a time-series-aware split (no shuffling), and performance is measured using precision, recall, and F1-score.

🔮 Prediction Visualization

The most recent known price is extended with a forecasted price point for July 2025, based on the predicted trend class:

• Drop → -5%
• Stable → ±0%
• Rise → +5%

This predicted point is visualized on a line chart, preserving historical context and highlighting known conflict months.

The model predicts price to be at 78.62, it currently sits at 74 (19 June 2025) . I will get back to it to see the accuracy