When samples from different classes over classification boundary are seriously overlapped, cluster assumption may not well reflect the real data distribution, so that semi-supervised classification methods based cluster assumption may yield even worse performance than their supervised counterparts. For the above unsafe semi-supervised classification problem, an Adjusted Cluster Assumption and Pairwise Constraints Jointly based Semi-Supervised Support Vector Machine classification method (ACA-JPC-S3VM) was proposed. On the one hand, the distances of individual unlabeled instances to the distribution boundary were considered in learning, which alleviated the degradation of the algorithm performance in such cases to some extent. On the other hand, the information of pairwise constraints was introduced to the algorithm to make up for its insufficient use of supervision information. The experimental results on the UCI dataset show that the performance of ACA-JPC-S3VM method would never be lower than that of SVM (Support Vector Machine), and the average accuracy is 5 percentage points higher than that of SVM when the number of labeled samples is 10. The experimental results on the image classification dataset show that the semi-supervised classification methods such as TSVM (Transductive SVM) have different degrees of unsafety learning (similar or worse performance than SVM) while ACA-JPC-S3VM can learn safely. Therefore, ACA-JPC-S3VM has better safety and correctness.

Conventional soft partition based clustering algorithms usually cannot achieve desired clustering outcomes in the situations where the data are quite spare or distorted. To address this problem, based on maximum entropy clustering, by means of the strategy of historical knowledge learning, two novel soft partition based clustering models called SPBC-RHK-1 and SPBC-RHK-2 for short respectively were proposed. SPBC-RHK-1 is the basic model which only refers to the historical cluster centroids, whereas SPBC-RHK-2 is of advanced modality based on the combination of historical cluster centroids and historical memberships. In terms of the historical knowledge, the effectiveness of both algorithms was improved distinctly, and SPBC-RHK-2 method showed better effectiveness and robustness compared to the other method since its higher ability of utilizing knowledge. In addition, because the involved historical knowledge does not expose the historical raw data, both of these two approaches have good capacities of privacy protection for historical data. Finally, experiments were conducted on both artificial and real-world datasets to verify above merits.