PranavSharma/dynamic-pricing-model

Model description

This is a regression model trained on the Dynamic Pricing Dataset. It was optimized using grid search with multiple hyperparameters.

Intended uses & limitations

This regression model is designed to predict the cost of rides based on various features such as expected ride duration, number of drivers, and time of booking.

Intended Uses:

• Dynamic Pricing Analysis: Helps optimize pricing strategies for ride-hailing platforms.
• Demand Forecasting: Supports business decisions by estimating cost trends based on ride-specific parameters.

Limitations:

• Feature Dependence: The model's accuracy is highly dependent on the input features provided.
• Dataset Specificity: Performance may degrade if applied to datasets with significantly different distributions.
• Outlier Sensitivity: Predictions can be affected by extreme values in the dataset.

Training Procedure

The model was trained using grid search to optimize hyperparameters. Cross-validation (5-fold) was performed to ensure robust evaluation. The best model was selected based on the lowest Mean Absolute Error (MAE) on the validation set.

Hyperparameters

Click to expand

Hyperparameter	Value
alpha	1
copy_X	True
fit_intercept	False
max_iter	1000
positive	False
precompute	False
random_state
selection	cyclic
tol	0.0001
warm_start	False

Model Plot

Lasso(alpha=1, fit_intercept=False)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Evaluation Results

The model achieved the following results on the test set:

• Mean Absolute Error (MAE): 50.31928636001356
• R² Score: 0.8854065597299239

Key Insights:

• Longer ride durations increase costs significantly, which may justify adding a surcharge for long-distance rides.
• Evening bookings reduce costs, potentially indicating lower demand during these hours.
• The model's accuracy is dependent on high-quality feature data.

Refer to the plots and tables for detailed performance insights.

Model Coefficients

Feature	Coefficient
Number_of_Riders	-0.1398
Number_of_Drivers	0.4665
Number_of_Past_Rides	-0.0033
Average_Ratings	-0.0000
Expected_Ride_Duration	3.4973
Location_Category_Suburban	0.0000
Location_Category_Urban	-0.0000
Customer_Loyalty_Status_Regular	0.0000
Customer_Loyalty_Status_Silver	0.0000
Time_of_Booking_Evening	-2.4212
Time_of_Booking_Morning	-0.0000
Time_of_Booking_Night	0.0000
Vehicle_Type_Premium	39.5754

Regression Equation

Cost of Ride = -0.1398 × Number_of_Riders + 0.4665 × Number_of_Drivers + -0.0033 × Number_of_Past_Rides + 3.4973 × Expected_Ride_Duration + -2.4212 × Time_of_Booking_Evening + 39.5754 × Vehicle_Type_Premium

Actual vs Predicted

The following plot shows the relationship between the actual and predicted values. The closer the points are to the diagonal line, the better the predictions. The dashed line represents the ideal case where predictions perfectly match the actual values.

The scatter plot above shows the predicted values against the actual values. The dashed line represents the ideal predictions where the predicted values are equal to the actual values.

How to Get Started with the Model

To use this model:

1. Install Dependencies: Ensure scikit-learn and pandas are installed in your environment.
2. Load the Model: Download the saved model file and load it using joblib:

   from joblib import load
   model = load('best_model.joblib')
   

3. Prepare Input Features: Create a DataFrame with the required input features in the same format as the training dataset.
4. Make Predictions: Use the predict method to generate predictions:

   predictions = model.predict(input_features)
   

Model Card Authors

This model card was written by Pranav Sharma.

Model Card Contact

For inquiries or feedback, you can contact the author via GitHub.

Citation

If you use this model, please cite it as follows:

@model{pranav_sharma_dynamic_pricing_model_2025,
  author       = {Pranav Sharma},
  title        = {Dynamic Pricing Model},
  year         = {2025},
  version      = {1.0.0},
  url          = {https://huggingface.co/PranavSharma/dynamic-pricing-model}
}