The prevalence of networked sensors and actuators in many real-world systems such as smart buildings, factories, power plants, and data centres generate substantial amounts of multivariate time series data for these systems. Many of these cyber-physical systems (CPSs) are engineered for mission-critical tasks and are thus targets for cyber-attacks. The rich sensor data can be continuously monitored for intrusion events through anomaly detection. However, conventional threshold-based anomaly detection methods are inadequate due to the dynamic complexities of these systems, while supervised machine learning methods are unable to exploit the large amounts of data due to the lack of labeled data. On the other hand, current unsupervised machine learning approaches have not fully exploited the spatial-temporal correlation and other dependencies amongst the multiple variables (sensors/actuators) in the system for detecting anomalies. Most of the current techniques also employed simple comparison between the present states and predicted normal ranges for anomaly detection, which can be inadequate given the highly dynamic behaviors of the systems. In this work, we propose an unsupervised multivariate anomaly detection method based on Generative Adversarial Networks (GANs), using the Long-Short-Term-Memory Recurrent Neural Networks (LSTM-RNN) as the base models (namely, the generator and discriminator) in the GAN framework to capture the temporal correlation of time series distributions. Instead of treating each data stream independently, our proposed Multivariate Anomaly Detection with GAN (MAD-GAN) framework considers the entire variable set concurrently to capture the latent interactions amongst the variables. We also fully exploit both the generator and discriminator produced by the GAN, using a novel anomaly score called DR-score to detect anomalies by discrimination and reconstruction. We have tested our proposed MAD-GAN using two recent datasets collected from real world CPS: the Secure Water Treatment (SWaT) and the Water Distribution (WADI) datasets. Our experimental results showed that the proposed MAD-GAN is effective in reporting anomalies caused by various cyberintrusions compared in these complex real-world systems.

The daily renewal of the corpus epithelium is fuelled by adult stem cells residing within tubular glands, but the identity of these stem cells remains controversial. Lgr5 marks homeostatic stem cells and 'reserve' stem cells in multiple tissues. Here, we report Lgr5 expression in a subpopulation of chief cells in mouse and human corpus glands. Using a non-variegated Lgr5-2A-CreERT2 mouse model, we show by lineage tracing that Lgr5-expressing chief cells do not behave as corpus stem cells during homeostasis, but are recruited to function as stem cells to effect epithelial renewal following injury by activating Wnt signalling. Ablation of Lgr5 cells severely impairs epithelial homeostasis in the corpus, indicating an essential role for these Lgr5 cells in maintaining the homeostatic stem cell pool. We additionally define Lgr5 chief cells as a major cell-of-origin of gastric cancer. These findings reveal clinically relevant insights into homeostasis, repair and cancer in the corpus.

No abstract

No abstract