A Deep Learning Approach To Multiple Kernel Fusion
This work addresses kernel fusion for activity recognition, but it appears incremental as it builds on existing deep learning and MKL methods.
The paper tackled the problem of multiple kernel fusion by proposing a deep neural network architecture that creates dense embeddings from kernel similarities and uses kernel dropout regularization with expanded composition kernels, achieving state-of-the-art performance on a real-world activity recognition dataset.
Kernel fusion is a popular and effective approach for combining multiple features that characterize different aspects of data. Traditional approaches for Multiple Kernel Learning (MKL) attempt to learn the parameters for combining the kernels through sophisticated optimization procedures. In this paper, we propose an alternative approach that creates dense embeddings for data using the kernel similarities and adopts a deep neural network architecture for fusing the embeddings. In order to improve the effectiveness of this network, we introduce the kernel dropout regularization strategy coupled with the use of an expanded set of composition kernels. Experiment results on a real-world activity recognition dataset show that the proposed architecture is effective in fusing kernels and achieves state-of-the-art performance.