Abstract
We introduce an automated Compositional Learning Framework, which can dynamically combine ML models to create a composite ML service. It leverages the MLOps principle to streamline drift-aware ML workflows. We showcase its applicability in the dy-namic Routing Modulation and Spectrum Allocation scenario with an open disaggregated control platform.
© 2024 The Author(s)
PDF ArticleMore Like This
Javier Errea, Huy Tran Quang, Dominique Verchere, Huu Trung Thieu, Andrea Mazzini, Lahcen Abnaou, Jelena Pesic, Marina Curtol, Abdelali El Imadi, Adlen Ksentini, and Djamal Zeghlache
M3Z.8 Optical Fiber Communication Conference (OFC) 2023
Filippo Cugini, Davide Scano, Andrea Sgambelluri, Francesco Paolucci, Alessio Giorgetti, and Piero Castoldi
W1C.4 Optical Fiber Communication Conference (OFC) 2024
Sai Chen, Weitang Zheng, Liang Dou, Huan Zhang, Zhao Sun, Lei Wang, Fan Gao, Boyuan Yan, Zhiqun Zhai, and Chongjin Xie
Th1G.6 Optical Fiber Communication Conference (OFC) 2024