@article {18535, title = {An Incremental Change Detection Test Based on Density Difference Estimation}, journal = {IEEE Transactions on Systems, Man, and Cybernetics: Systems}, volume = {47}, year = {2017}, month = {Oct}, pages = {2714-2726}, keywords = {Big data framework, Change detection, change detection method, density difference estimation, Estimation, false positive rates, Feature extraction, Histograms, incremental change detection test, incremental computing, incremental least squares density difference change detection method (LSDD-Inc), Internet of Things, Kernel, least squares approximations, least squares density difference, linear combination, LSDD, LSDD values, nonoverlapping data windows, probability, Probability density function, probability density function (pdf)-free, probability density functions, process continuous data streams, Training}, issn = {2168-2216}, doi = {10.1109/TSMC.2017.2682502}, author = {Bu, Li and Zhao, Dongbin and Alippi, Cesare} } @article {18540, title = {A Kolmogorov-Smirnov Test to Detect Changes in Stationarity in Big Data}, journal = {IFAC-PapersOnLine}, volume = {50}, year = {2017}, pages = {14260 - 14265}, abstract = {The paper proposes an effective change detection test for online monitoring data streams by inspecting the least squares density difference (LSDD) features extracted from two non-overlapped windows. The first window contains samples associated with the pre-change probability distribution function (pdf) and the second one with the post-change one (that differs from the former if a change in stationarity occurs). This method can detect changes by also controlling the false positive rate. However, since the window sizes is fixed after the test has been configured (it has to be small to reduce the execution time), the method may fail to detect changes with small magnitude which need more samples to reach the requested level of confidence. In this paper, we extend our work to the Big Data framework by applying the Kolmogorov-Smirnov test (KS test) to infer changes. Experiments show that the proposed method is effective in detecting changes.}, keywords = {change detection test, KS test, LSDD}, issn = {2405-8963}, doi = {https://doi.org/10.1016/j.ifacol.2017.08.1821}, url = {http://www.sciencedirect.com/science/article/pii/S2405896317324436}, author = {Zhao, Dongbin and Bu, Li and Alippi, Cesare and Wei, Qinglai} } @conference {18449, title = {Ensemble LSDD-based Change Detection Tests}, booktitle = {IEEE-INNS International Joint Conference on Neural Networks (IJCNN16)}, year = {2016}, month = {07/2016}, address = {Vancouver, Canada}, abstract = {The least squares density difference change detection test (LSDD-CDT) has proven to be an effective method in detecting concept drift by inspecting features derived from the discrepancy between two probability density functions (pdfs). The first pdf is associated with the concept drift free case, the second to the possible post change one. Interestingly, the method permits to control the ratio of false positives. This paper introduces and investigates the performance of a family of LSDD methods constructed by exploring different ensemble options applied to the basic CDT procedure. Experiments show that most of proposed methods are characterized by improved performance in change detection once compared with the direct ensemble-free counterpart.}, author = {Bu, Li and Alippi, Cesare and Zhao, Dongbin} }