Over ons

Consortium

Team

AI Vlaanderen

Contact

<h5>Description:</h5>
<p>An MCMC-based factorization method Macau that is able to scale to millions of features, Fu &gt; 1,000,000 while the factorized matrix Y can have a million of rows (and/or columns). In contrast, the sampling-based Bayesian factorization methods Bayesian Matrix Factorization with Side Information (BMFSI) &nbsp;and Bayesian Canonical PARAFAC with Features and Networks (BCPFN) have complexity O(Fu^3) because they require explicit computation and Cholesky decomposition of the covariance matrix of size Fu&times;Fu for linking the side information to the factorization.</p>

<h5>More Information:</h5>
<ul>
<li><strong>Grand Challenge/Workpackage(s)</strong>: <a href="/nl/onderzoek/onderzoeksuitdagingen/ai-driven-data-science" data-entry-id="58" target="_blank" rel="noopener" data-entry-id="58">GC1</a> - WP5</li>
<li><strong>Responsible Research Lead</strong>: <a href="mailto:yves.moreau@kuleuven.be" target="_blank" rel="noopener">Yves Moreau&nbsp;</a></li>
<li><strong>Authors</strong>: <a href="mailto:jaak.simm@esat.kuleuven.be" target="_blank" rel="noopener">Jaak Simm</a>, <a href="mailto:adam.arany@esat.kuleuven.be" target="_blank" rel="noopener">Adam Arany</a>, <a href="mailto:yves.moreau@kuleuven.be" target="_blank" rel="noopener">Yves Moreau</a></li>
<li><strong>Programming Language</strong>: Python/Cython/C++</li>
<li><a href="https://macau.readthedocs.io/en/latest/source/examples.html" target="_blank" rel="noopener">Link to the website</a></li>
<li><a href="https://github.com/jaak-s/macau" target="_blank" rel="noopener">Link to the repository</a></li>
<li><strong>Categories:</strong> Supervised learning, &nbsp;Recommender systems</li>
</ul>

<h5>Publications:</h5>
<ul>
<li>J. Simm, A. Arany, P. Zakeri, T. Haber,&nbsp; J. K. Wegner, V. Chupakhin, H. Ceulemans, Y. Moreau, "<a href="https://ieeexplore.ieee.org/document/8168143" target="_blank" rel="noopener">Macau: Scalable Bayesian factorization with high-dimensional side information using MCMC</a>", In proceedings of the 2017 IEEE 27th International Workshop on Machine Learning for Signal Processing (MLSP), 2017</li>
</ul>

<p>Scalable Bayesian factorization with high-dimensional side information using MCMC</p>