Deep subspace clustering network has shown its effectiveness in hyperspectral image (HSI) clustering. However, there are two major challenges that need to be addressed: 1) lack of effective supervision for feature learning; 2) negative effect caused by the high redundancy of the global dictionary atoms. In this paper, we propose an end-to-end trainable network for HSI clustering. Specifically, to ensure the extracted features are wellsuited to subsequent subspace clustering, the cluster assignments with high confidence are employed as pseudo-labels to supervise the feature learning process. Then, an adaptive self-expressive coefficient matrix initialization strategy is designed to reduce the dictionary redundancy, where the spectral similarity between each target sample and its neighbors are modeled via the knearest neighbor graph to guide the initialization. Experimental results on three public HSI datasets demonstrate the effectiveness of the proposed method. In particular, our method outperforms several state-of-the-art HSI clustering methods, and achieves overall accuracy (OA) of 100% on both SalinasA and Pavia University datasets.