Netajam/Decision-Optimizer

Decision optimizer is a Python library that facilitates decision-making based on the probability of events and their associated utility. It provides a comprehensive toolkit for defining probabilistic events, combining them into outcomes, and evaluating decisions based on expected utility.

Decision-Optimizer based on Event Probabilities and Utility

This repository provides a Python library for making decisions based on the probability of events and their associated utility. It allows users to define probabilistic events, combine them into outcomes, and evaluate decisions based on the expected utility.

Installation

1. Clone the repository:

   git clone https://github.com/Netajam/Decision-Optimizer.git
   

2. Install the required dependencies:

   pip install -r requirements.txt   
   

Usage

1. Import the necessary modules from the src package:

   from src.probabilistic_events import Event
   from src.visualization import plot_distribution, plot_utility_distribution, plot_utility_distribution_decision, plot_probability_distribution
   from src.decisions import Decision
   from src.outcomes import Outcome
   from src.decision_evaluation import evaluate_decision

2. Define probabilistic events using the Event class:

   E1 = Event("E1", "normal", {'mean': 0.3, 'std': 0.1})

3. Create decisions using the Decision class:

   decision1 = Decision("Buying my train ticket")

4. Define outcomes by combining probabilistic events and specifying utility functions:

   outcome1 = Outcome("E1&E2&E3", decision1, [E1, E2, E3], combine_formula, utility_function)

5. Evaluate decisions based on the outcomes:

   final_utilities, weighted_average_utility, all_utility_samples = evaluate_decision(decision1, all_outcomes)

6. Visualize the probability distributions, utility distributions, and decision evaluations using the provided visualization functions.

For more detailed examples, please refer to the Jupyter notebooks in the notebooks directory.

Contributing

Contributions are welcome! If you find any issues or have suggestions for improvements, please open an issue or submit a pull request.

License

This project is licensed under the MIT License.