Electroencephalogram (EEG) contains important physiological information that can reflect the activity of human brain, making it useful for epileptic seizure detection and epilepsy diagnosis. However visual inspection of large amounts of EEG by human expert is time-consuming, and meanwhile there are often inconsistencies in judgement between physicians. In this paper, we develop a unified framework for early epileptic seizure detection and epilepsy diagnosis, which includes two phases. In the first phase, the signal intensity is first calculated for each data point of the given EEG, enabling the well-known autoregressive moving average (ARMA) model to characterize the dynamic behavior of the EEG time series. The residual error between the predicted value of learned ARMA model and the actually observed value is used as the anomaly score to support a null hypothesis testing for making epileptic seizure decision. The epileptic seizure detection phase can provide a quick detection for anomaly EEG patterns, but the resulting suspicious segment may include epilepsy or other disordering EEG activities thus required to be identified. Therefore, in the second phase, we use pattern recognition technique to classify the suspicious EEG segments. In particular, we propose a new and practical classifier based on a pairwise of one-class SVMs for epilepsy diagnosis. The proposed classifier requires normal and epilepsy data for training, but can recognize normal, epilepsy and even other disorders that would not be trained in the training samples. This point is practical and meaningful in real clinic scenarios as the input EEG may include other brain disordering diseases besides of epilepsy. We conducted experiments on the publicly-available Bern-Barcelona and CHB-MIT EEG database, respectively, to validate the effectiveness of the proposed framework, and our method achieved classification accuracy of 93% and 94% on them. Comprehensive experimental results, outperforming the state-of-...

This paper suggests the correlation information, hidden in spatial configuration and temporal dynamic of frequencies, is an important indication for bearing health condition. To consider this information, we extend graph-modeling strategy, and introduce a bearing fault detection and diagnosis technique based on temporal-spatio graph. First, short-time periodogram is extracted from vibration signal, and then modeled by a temporal-spatio graph. In fault detection phase, the spectrum of temporal channel graph is used to map short-time periodogram to acquire the so-called graph-mapped spectrum (GMS). The principal frequency in resulting GMS is found highly related with the health condition of monitored bearing. Thus, any change of health condition can be detected by checking this principal frequency over time. Once a fault is detected, the spatio channel graph is fed to K-Nearest Neighbor classifier, coupled with a specific graph distance metric, for fault type identification. Comprehensive experiments on two benchmarking data sets along with theoretical interpretation demonstrate the superiority of proposed method over state-of-the-arts. The proposed temporal-spatio graph provides a signif- icant extension of existing spectrum analysis for fault detection and diagnosis.

Code summarization has long been viewed as a challenge in software engineering because of the difficulties of understanding source code and generating natural language. Some mainstream methods combine abstract syntax trees with language models to capture the structural information of the source code and generate relatively satisfactory comments. However, these methods are still deficient in code understanding and limited by the long dependency problem. In this paper, we propose a novel model called Fret, which stands for Functional REinforced Transformer with BERT. The model provides a new way to generate code comments by learning code functionalities and deepening code understanding while alleviating the problem of long dependency. For this purpose, a novel reinforcer is proposed for learning the functional contents of code so that more accurate summaries to describe the code functionalities can be generated. In addition, a more efficient algorithm is newly designed to capture the source code structure. The experimental results show that the effectiveness of our model is remarkable. Fret significantly outperforms all the state-of-the-art methods we examine. It pushes the BLEU-4 score to 24.32 for Java code summarization (14.23% absolute improvement) and the ROUGE-L score to 40.12 for Python. An ablation test is also conducted to further explore the impact of each component of our method.

Detection of structural changes in machine state attracts increasing attention for the monitoring and prognosis of rotating machinery. Very recently, the graph model has been introduced and adopted for modeling of normal machine state so as to investigate the likelihood of potential changes by the use of Martingale-test. This paper expands the potential of the graph model toward correlation-analysis-based monitoring applications, where the problem of measuring graph similarity is the main challenge due to domain specificity. We first investigated six typical schemes taken from other areas for this purpose, and found that they show discriminative capacities when dealing with changes with different types in the machine condition monitoring scenarios. Subsequently, based on a procedure called “method of ranks,” we have chosen four schemes among them which are potentially promising for our usage. Meanwhile, based on these metrics, two machine learning based similarity metrics are proposed to further improve their practical values by combining them. At last, comprehensive theoretical interpretations and comparisons of presented methods are made both in simulated scenarios and real-engineering monitoring applications.