A Bayesian Hidden Markov Model-based approach for anomaly detection in electronic systems

Dorj, Enkhjargal; Chen, Chaochao; Pecht, Michael

doi:10.1109/aero.2013.6497204

Cited by 23 publications

(10 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The disadvantages of Euclidean distance lie in the way of calculating the mean and standard deviation of the data and the existence of correlation between the axes, which makes the feature vector difficult. In the study by Dorj et al, 24 a feature extraction method for fault diagnosis is presented. In this work, the hidden Markov model classification technique is used to detect a fault in the system.…”

Section: Introductionmentioning

confidence: 99%

Fault isolation of analog circuit using an optimized ensemble empirical mode decomposition approach based on multi-objective optimization

Moezi

Kargar

2021

Proceedings of the Institution of Mechanical Engineers, Part I:

View full text Add to dashboard Cite

This article proposed a practical approach to isolating faults in analog circuits. The contribution of this article is twofold. First, the optimized empirical mode decomposition approach is presented based on the Hellinger distance such that there is a minimum dependency between intrinsic mode functions. Features with high distinction could be extracted by employing intrinsic mode functions in fault detection problem of analog benchmark circuits. Second, the non-dominated sorting genetic algorithm is employed to retain excellent features and speed up the execution, resulting in the high accuracy of fault detection and isolation. The number of features and mean squared error are selected as objective functions. The features from the data are also extracted using the fast Fourier and wavelet transforms for comparison. Finally, the support vector machine and artificial neural network are employed to isolate faults. Two circuits under test are simulated, and the output signals of the faulty and fault-free circuits are extracted by the Monte Carlo analysis. According to the obtained simulation results, the proposed method with a low-dimensional feature vector outperformed the previous methods, and the computational time has also reduced significantly.

show abstract

Section: Introductionmentioning

confidence: 99%

Fault isolation of analog circuit using an optimized ensemble empirical mode decomposition approach based on multi-objective optimization

Moezi

Kargar

2021

Proceedings of the Institution of Mechanical Engineers, Part I:

View full text Add to dashboard Cite

show abstract

“…This method can be found for example in (Chandola et al, 2012), (Lane and Brodley, 2003), (Joshi and Phoha, 2005), (Khreich et al, 2009), (Khreich et al, 2010), (Lane and Brodley, 2003) or (Wang et al, 2004). A similar idea can be found in (Dorj et al, 2013), which replaces HMMs with Bayesian HMMs. The problem with all those techniques is that they are not very good in modeling multiple types of sequences.…”

Section: Related Workmentioning

confidence: 99%

Anomaly Detection with HMM Gauge Likelihood Analysis

Lorbeer

Deutsch

Ruppel

et al. 2019

2019 IEEE Fifth International Conference on Big Data Computing Service and Applications (BigDataService)

View full text Add to dashboard Cite

This paper describes a new method, HMM gauge likelihood analysis, or GLA, of detecting anomalies in discrete time series using Hidden Markov Models and clustering. At the center of the method lies the comparison of subsequences. To achieve this, they first get assigned to their Hidden Markov Models using the Baum-Welch algorithm. Next, those models are described by an approximating representation of the probability distributions they define. Finally, this representation is then analyzed with the help of some clustering technique or other outlier detection tool and anomalies are detected. Clearly, HMMs could be substituted by some other appropriate model, e.g. some other dynamic Bayesian network. Our learning algorithm is unsupervised, so it doesn't require the labeling of large amounts of data. The usability of this method is demonstrated by applying it to synthetic and real-world syslog data.

show abstract

“…Applications of HMMs appear in various fields. Dorj et al (2013), as an example, proposed a data-driven approach for anomaly detection in electronic systems based on a Bayesian Hidden Markov model classification technique. Shi and Sun (2012) studied on the HMM model based on system calls anomaly detection in order to improve the detection accuracy.…”

Section: Hidden Markov Models (Hmms)mentioning

confidence: 99%

Hidden Markov Models With Covariates for Analysis of Defective Industrial Machine Parts

Sirima

Pokorný

2014

Journal of Mathematics and Statistics

View full text Add to dashboard Cite

Monthly counts of industrial machine part errors are modeled using a two-state Hidden Markov Model (HMM) in order to describe the effect of machine part error correction and the amount of time spent on the error correction on the likelihood of the machine part to be in a "defective" or "non-defective" state. The number of machine parts errors were collected from a thermo plastic injection molding machine in a car bumper auto parts manufacturer in Liberec city, Czech Republic from January 2012 to November 2012. A Bayesian method is used for parameter estimation. The results of this study indicate that the machine part error correction and the amount of time spent on the error correction do not improve the machine part status of the individual part, but there is a very strong month-to-month dependence of the machine part states. Using the Mean Absolute Error (MAE) criterion, the performance of the proposed model (MAE = 1.62) and the HMM including machine part error correction only (MAE = 1.68), from our previous study, is not significantly different. However, the proposed model has more advantage in the fact that the machine part state can be explained by both the machine part error correction and the amount of time spent on the error correction.

show abstract

A Bayesian Hidden Markov Model-based approach for anomaly detection in electronic systems

Cited by 23 publications

References 17 publications

Fault isolation of analog circuit using an optimized ensemble empirical mode decomposition approach based on multi-objective optimization

Fault isolation of analog circuit using an optimized ensemble empirical mode decomposition approach based on multi-objective optimization

Anomaly Detection with HMM Gauge Likelihood Analysis

Hidden Markov Models With Covariates for Analysis of Defective Industrial Machine Parts

Contact Info

Product

Resources

About