科研成果

In this paper, a group-sensitive multiple kernel learning (GS-MKL) method is proposed for object recognition to accommodate the intra-class diversity and the inter-class correlation. By introducing the group between the object category and individual images as an intermediate representation, GS-MKL attempts to learn group-sensitive multi-kernel combinations together with the associated classifier. For each object category, the image corpus from the same category is partitioned into groups. Images with similar appearance are partitioned into the same group, which corresponds to the sub-category of the object category. Accordingly, intra-class diversity can be represented by the set of groups from the same category but with diverse appearances; Inter-class correlation can be represented by the correlation between the groups from different categories. GS-MKL provides a tractable solution to adapt multikernel combination to local data distribution and to seek a trade-off between capturing the diversity and keeping the invariance for each object category. Different from the simple hybrid grouping strategy that solves sample grouping and GS-MKL training independently, two sample grouping strategies are proposed to integrate sample grouping and GS-MKL training. The first one is looping hybrid grouping method where global kernel clustering method and GS-MKL interact with each other by sharing group-sensitive multi- kernel combination. The second one is dynamic divisive grouping method where hierarchical kernel-based grouping process interacts with GS-MKL. Experimental results show that performance of GS-MKL does not vary significantly with different grouping strategies, but looping hybrid grouping method produces slightly better results. On four challenging datasets, our proposed method has achieved encouraging performance comparable to the state-of-the-art and outperformed several existing MKL methods.

How can one country narrow the regional disparity during the tremendous expansion of higher education? This issue remains unexamined and critical analysis is needed to unveil the spatial dynamics behind expansion of higher education. The spatial analysis shows that there is significant strategic interaction among neighboring provinces in China during the expansion of higher education. It implies that the expansion of higher education intensifies spatial interaction among provinces which facilitates narrowing gap of higher education among provinces in China. The convergence analysis shows that the speed of σ-convergence is 1.3% and the average speed of β-convergence is 2.1%, which give robust evidences to support the increasing spatial equalization of China’s higher education. Meanwhile, local spatial dynamics analysis shows that the catch-up strategy and cluster growth strategies of China are effective to achieve the balanced expansion goal.