MKLpy: a python-based framework for Multiple Kernel Learning
This work offers a tool for researchers and practitioners in machine learning to more easily implement and experiment with Multiple Kernel Learning methods, but it is incremental as it focuses on software development rather than novel algorithmic advances.
The authors introduced MKLpy, a Python-based framework for Multiple Kernel Learning, which provides algorithms for classification, kernel computation mechanisms for various data types, and evaluation strategies to enhance usability and simplify development.
Multiple Kernel Learning is a recent and powerful paradigm to learn the kernel function from data. In this paper, we introduce MKLpy, a python-based framework for Multiple Kernel Learning. The library provides Multiple Kernel Learning algorithms for classification tasks, mechanisms to compute kernel functions for different data types, and evaluation strategies. The library is meant to maximize the usability and to simplify the development of novel solutions.