This is the accompanying code for the paper
Hyper-Parameter Optimization for Latent Spaces in Dynamic Recommender Systems, by Bruno Veloso, Luciano Caroprese, Matthias König, Sónia Teixeira, Giuseppe Manco, Holger H. Hoos, and João Gama. The paper has been accepted for presentation as a full paper at the research track of the 2021 edition of The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases.
The paper is the result of a joint activity at the HumaneAI-Net project, as described in the related page. A poster describing the main results achieved within this activity are described here.
The following video exemplifies the behavior of the Nelder-Mead algorithm proposed in the paper.
The notebook is essentially self-explanatory. You need Pytorch >= 1.5 to run it.
- There is a section containing the data generator proposed in the paper.
- Experiments including the Nelder-Mead, SMAC baseline and Static baseline are reported and all the graphs contained in the paper are reproduced.
Bibtex
@inproceedings{VelEtAl21,
title="Hyper-Parameter Optimization for Latent Spaces in Dynamic Recommender Systems",
author="Veloso, Bruno and Caroprese, Luciano and K{\"o}nig, Matthias and Teixeira, S{\'o}nia and Manco, Giuseppe and Hoos, Holger H and Gama, Jo{\~a}o",
booktitle="Proceedings of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases",
year="2021"
tags="AutoML, Hyper-Parameter Optimization, Latent Spaces, Nelder-Mead, SMAC, Recommender Systems"
}