idinn 0.1.2

Inventory dynamics–informed neural networks for solving single-sourcing and dual-sourcing problems.

idinn: Inventory-Dynamics Control with Neural Networks

idinn implements inventory dynamics–informed neural networks for solving single-sourcing and dual-sourcing problems. Neural network controllers and inventory dynamics are implemented into easily customizable classes to enable users to find the optimal controllers for the user-specified inventory systems.

Requirements

The basic usage of idinn requires working Python and PyTorch installation. If plotting simulation result of a controller is needed, matplotlib should also be installed.

Installation

The package can be installed form the git repository. To do that, run

python -m pip install git+https://gitlab.com/ComputationalScience/inventory-optimization.git@main

Or, if you want to inspect the source code and edit locally, run

git clone https://gitlab.com/ComputationalScience/inventory-optimization.git
cd inventory-optimization
python -m pip install -e .

Example Usage

import torch
from idinn.sourcing_model import SingleSourcingModel
from idinn.controller import SingleFullyConnectedNeuralController

# Initialize the sourcing model and the neural controller
sourcing_model = SingleSourcingModel(
    lead_time=0, holding_cost=5, shortage_cost=495, batch_size=32, init_inventory=10
)
controller = SingleFullyConnectedNeuralController(
    hidden_layers=[2], activation=torch.nn.CELU(alpha=1)
)
# Train the neural controller
controller.train(
    sourcing_model=sourcing_model,
    sourcing_periods=50,
    validation_sourcing_periods=1000,
    epochs=5000,
    tensorboard_writer=torch.utils.tensorboard.SummaryWriter(),
    seed=1,
)
# Simulate and plot the results
controller.plot(sourcing_model=sourcing_model, sourcing_periods=100)
# Calculate the optimal order quantity for applications
controller.forward(
    current_inventory=torch.tensor([[10]]),
    past_orders=torch.tensor([[1, 5]]),
)

Documentation

For tutorials and documentation, please refer to our documentation.

Papers using idinn

We will add papers that use ìdinn to this list as they appear online.

• Böttcher, Lucas, Thomas Asikis, and Ioannis Fragkos. "Control of Dual-Sourcing Inventory Systems Using Recurrent Neural Networks." INFORMS Journal on Computing 35.6 (2023): 1308-1328.

Contributors

• Jiawei Li
• Thomas Asikis
• Ioannis Fragkos
• Lucas Böttcher