Online Inventory Problems: Beyond the i.i.d. Setting with Online Convex Optimization

Authors: Massil HIHAT, Stéphane Gaïffas, Guillaume Garrigos, Simon Bussy

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, in Section 6 we present numerical experiments on both synthetic and real-world data that validate empirically the versatility and performances of Max COSD.
Researcher Affiliation Collaboration 1LOPF, Califrais Machine Learning Lab, Paris, France 2Université Paris Cité and Sorbonne Université, CNRS, Laboratoire de Probabilités, Statistique et Modélisation, Paris, France
Pseudocode Yes Algorithm 2: Max COSD
Open Source Code Yes The code is available at https://github.com/Califrais/newsvendor_tester.
Open Datasets Yes Demands are taken from the real-world dataset of the M5 competition [17].
Dataset Splits No The paper mentions using the M5 competition dataset for T = 1969 periods but does not provide specific details on training, validation, or test splits.
Hardware Specification No The paper does not provide any specific hardware details such as GPU or CPU models, or memory specifications used for running the experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python version, library versions like PyTorch or scikit-learn).
Experiment Setup Yes All the algorithms have been initialized with y1 = 0. Settings 1, 2 and 3 have been run 10 times, with different demand realizations generated through independent samples. Figure 1 shows, for every setting, the regret obtained after T periods as a function of the learning rate parameter γ [10−5, 101]. We picked T = 1969 for all the settings...