Time series motifs are approximately repeated patterns found within the data. Such motifs have utility for many data mining algorithms, including rule-discovery, novelty-detection...
Dragomir Yankov, Eamonn J. Keogh, Jose Medina, Bil...
The problem of identifying mislabeled training examples has been examined in several studies, with a variety of approaches developed for editing the training data to obtain better...
We consider the task of performing anomaly detection in highly noisy multivariate data. In many applications involving real-valued time-series data, such as physical sensor data a...
Most malicious attacks compromise system security through memory corruption exploits. Recently proposed techniques attempt to defeat these attacks by protecting program control da...
Shuo Chen, Jun Xu, Nithin Nakka, Zbigniew Kalbarcz...
Probabilistic mixture models are used for a broad range of data analysis tasks such as clustering, classification, predictive modeling, etc. Due to their inherent probabilistic na...