Outlier detection for process control data based on a non-linear Auto-Regression Hidden Markov Model method

2011 ◽  
Vol 34 (5) ◽  
pp. 527-538 ◽  
Author(s):  
Fang Liu ◽  
Zhizhong Mao ◽  
Weixing Su
Keyword(s):  
Process Control ◽  
Markov Model ◽  
Hidden Markov Model ◽  
Outlier Detection ◽  
Hidden Markov ◽  
Control Data ◽  
Model Method ◽  
Non Linear ◽  
Auto Regression

scholarly journals A Hybrid Hidden Markov Model for Pipeline Leakage Detection

2021 ◽  
Vol 11 (7) ◽  
pp. 3138
Author(s):  
Mingchi Zhang ◽  
Xuemin Chen ◽  
Wei Li
Keyword(s):  
Neural Network ◽  
Markov Model ◽  
Hidden Markov Model ◽  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Deep Neural Network ◽  
Hidden Markov ◽  
Gaussian Mixture ◽  
State Sequence ◽  
Non Linear

In this paper, a deep neural network hidden Markov model (DNN-HMM) is proposed to detect pipeline leakage location. A long pipeline is divided into several sections and the leakage occurs in different section that is defined as different state of hidden Markov model (HMM). The hybrid HMM, i.e., DNN-HMM, consists of a deep neural network (DNN) with multiple layers to exploit the non-linear data. The DNN is initialized by using a deep belief network (DBN). The DBN is a pre-trained model built by stacking top-down restricted Boltzmann machines (RBM) that compute the emission probabilities for the HMM instead of Gaussian mixture model (GMM). Two comparative studies based on different numbers of states using Gaussian mixture model-hidden Markov model (GMM-HMM) and DNN-HMM are performed. The accuracy of the testing performance between detected state sequence and actual state sequence is measured by micro F1 score. The micro F1 score approaches 0.94 for GMM-HMM method and it is close to 0.95 for DNN-HMM method when the pipeline is divided into three sections. In the experiment that divides the pipeline as five sections, the micro F1 score for GMM-HMM is 0.69, while it approaches 0.96 with DNN-HMM method. The results demonstrate that the DNN-HMM can learn a better model of non-linear data and achieve better performance compared to GMM-HMM method.

scholarly journals A Proposed Hidden Markov Model Method for Dynamic Device Pairing on Internet of Things End-Devices

2021 ◽  
Vol 14 (3) ◽  
pp. 274-285
Author(s):  
Aji Gautama Putrada ◽  
Nur Ghaniaviyanto Ramadhan
Keyword(s):  
Markov Model ◽  
Hidden Markov Model ◽  
Hidden Markov ◽  
Access Point ◽  
Received Signal Strength Indicator ◽  
Model Method ◽  
Iot Platform ◽  
High Level ◽  
Device Pairing ◽  
Better Than

Dynamic device pairing is a context-based zero-interaction method to pair end-devices in an IoT System based on Received Signal Strength Indicator (RSSI) values. But if RSSI detection is done in high level, the accuracy is troublesome due to poor sampling rates. This research proposes the Hidden Markov Model method to increase the performance of dynamic device pairing detection. This research implements an IoT system consisting an Access Point, an IoT End Device, an IoT Platform, and an IoT application and performs a comparison of two different methods to prove the concept. The results show that the precision of dynamic device pairing with HMM is better than without HMM and the value is 83,93%.

scholarly journals An Enhanced Hidden Semi-Markov model for Outlier Detection in Multivariate Datasets

2021 ◽  
Author(s):  
G Manoharan ◽  
K Sivakumar
Keyword(s):  
Markov Model ◽  
Hidden Markov Model ◽  
Outlier Detection ◽  
Hidden Markov ◽  
Research Work ◽  
Detection Accuracy ◽  
Detection Techniques ◽  
Detection Model ◽  
Huge Data ◽  
Proposed Model

Outlier detection in data mining is an important arena where detection models are developed to discover the objects that do not confirm the expected behavior. The generation of huge data in real time applications makes the outlier detection process into more crucial and challenging. Traditional detection techniques based on mean and covariance are not suitable to handle large amount of data and the results are affected by outliers. So it is essential to develop an efficient outlier detection model to detect outliers in the large dataset. The objective of this research work is to develop an efficient outlier detection model for multivariate data employing the enhanced Hidden Semi-Markov Model (HSMM). It is an extension of conventional Hidden Markov Model (HMM) where the proposed model allows arbitrary time distribution in its states to detect outliers. Experimental results demonstrate the better performance of proposed model in terms of detection accuracy, detection rate. Compared to conventional Hidden Markov Model based outlier detection the detection accuracy of proposed model is obtained as 98.62% which is significantly better for large multivariate datasets.

scholarly journals Blind bandwidth extension of audio signals based on non-linear prediction and hidden Markov model

2014 ◽  
Vol 3 ◽  
Author(s):  
Xin Liu ◽  
Changchun Bao
Keyword(s):  
Markov Model ◽  
Hidden Markov Model ◽  
State Space ◽  
Linear Prediction ◽  
Hidden Markov ◽  
Audio Signals ◽  
Bandwidth Extension ◽  
Fine Spectrum ◽  
Non Linear ◽  
Audio Codec

The bandwidth limitation of wideband (WB) audio systems degrades the subjective quality and naturalness of audio signals. In this paper, a new method for blind bandwidth extension of WB audio signals is proposed based on non-linear prediction and hidden Markov model (HMM). The high-frequency (HF) components in the band of 7–14 kHz are artificially restored only from the low-frequency information of the WB audio. State-space reconstruction is used to convert the fine spectrum of WB audio to a multi-dimensional space, and a non-linear prediction based on nearest-neighbor mapping is employed in the state space to restore the fine spectrum of the HF components. The spectral envelope of the resulting HF components is estimated based on an HMM according to the features extracted from the WB audio. In addition, the proposed method and the reference methods are applied to the ITU-T G.722.1 WB audio codec for comparison with the ITU-T G.722.1C super WB audio codec. Objective quality evaluation results indicate that the proposed method is preferred over the reference bandwidth extension methods. Subjective listening results show that the proposed method has a comparable audio quality with G.722.1C and improves the extension performance compared with the reference methods.

scholarly journals Anomaly Detection in Control Data of Programming Execution by Context Sensitive Hidden Markov Model

2020 ◽  
Vol 9 (4) ◽  
pp. 1866-1871
Keyword(s):  
Machine Learning ◽  
Anomaly Detection ◽  
Markov Model ◽  
Hidden Markov Model ◽  
Hidden Markov ◽  
Detection System ◽  
Control Data ◽  
Context Sensitive ◽  
Wide Range ◽  
Artificial Learning

The inconsistency is a major problem in security of information in computer is two ways: data inconsistency and application inconsistency. These two problems are raised due to bad structure of design in programming and create security breaches, vulnerable entries by exploiting application codes. So we can discover these anomalies by design of anomaly detection system (ADS) models at system programming (coding) levels with the help of machine learning. The security vulnerabilities (anomalies) are frequently occurred at potential code execution by exploitation or manipulation of instructions. So, in this paper we have specified various forms of extensions to our work to detect wide range of anomalies at coding exploits and use of a machine learning technique called Context Sensitive-Hidden Markov Model (CS-HMM) will improve the overall performance of ADS by discovering the correlations between control data instances. In this paper we are going to use Linux OS tracing kits to collect the necessary information such as control data instances (return addresses) collected from system as part of artificial learning. The results evaluated through practice on various programs developed for work and also uses of some Linux commands for tracing, finally compared performance of all those input datasets generated live (artificially). After that, the CS-HMM is applying to datasets to scrutinize the anomalies with similarity-search and correlation of function control data of program and classification process determines the anomalous outcomes.

Sign in / Sign up

Export Citation Format

Share Document