Summary of Changes and Responses to Reviews

To make the problem of multi-label classification with many classes more tractable, in recent years academia has seen efforts devoted to performing label space dimension reduction (LSDR). Specifically, LSDR encodes high-dimensional label vectors into low-dimensional code vectors lying in a latent space, so as to train predictive models at much lower costs. With respect to the prediction, it performs classification for any unseen instance by recovering a label vector from its predicted code vector via a decoding process. In this paper, we propose a novel method, namely End-to-End Feature-aware label space Encoding (EFE), to perform LSDR. Instead of requiring an encoding function like most previous works, EFE directly learns a code matrix formed by code vectors of the training instances in an endto-end manner. Another distinct property of EFE is its feature awareness attributable to the fact that the code matrix is learnt by jointly maximizing the recoverability of the label space and the predictability of the latent space. Based on the learnt code matrix, EFE further trains predictive models to map instance features into code vectors, and also learns a linear decoding matrix for efficiently recovering the label vector of any unseen instance from its predicted code vector. Theoretical analyses show that both the code matrix and the linear decoding matrix in EFE can be efficiently learnt. Moreover, similar to previous works, EFE can be specified to learn an encoding function. And it can also be extended with kernel tricks to handle non-linear correlations between the feature space and the latent space. Comprehensive experiments conducted on diverse benchmark datasets with many classes show consistent performance gains of EFE over the stateof-the-art methods.