{"id":4338,"date":"2022-05-24T13:42:47","date_gmt":"2022-05-24T05:42:47","guid":{"rendered":"https:\/\/ascendas-asia.com\/?page_id=4338"},"modified":"2022-05-27T10:40:08","modified_gmt":"2022-05-27T02:40:08","slug":"anomaly-detection","status":"publish","type":"page","link":"https:\/\/ascendas-asia.com\/th\/resources\/anomaly-detection\/","title":{"rendered":"Anomaly Detection"},"content":{"rendered":"<h3 class=\"h3 add_font_color_orange\" style=\"text-align: justify;\"><span style=\"color: #e67e23;\"><strong>Identify unexpected events and departures from normal behavior<\/strong><\/span><\/h3>\n<div style=\"text-align: justify;\">\n<div class=\"mainParsys parsys containsResourceName resourceClass-parsys\">\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\">\n<p>Anomaly detection is the process of identifying events or patterns that differ from expected behavior. Anomaly detection can range from simple outlier detection to complex machine learning algorithms trained to uncover hidden patterns across hundreds of signals. Engineers and data scientists use anomaly detection to identify:<\/p>\n<ul>\n<li>Faults in machinery for predictive maintenance<\/li>\n<li>Defects in manufacturing production lines<\/li>\n<li>Cancer in radiology images<\/li>\n<li>Fraud in financial transactions<\/li>\n<li>Customer churn in retail<\/li>\n<li>Unusual movements in video surveillance footage<\/li>\n<\/ul>\n<\/div>\n<\/div>\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\">\n<p>There are many ways to design anomaly detection algorithms in MATLAB<sup>\u00ae<\/sup>. The anomaly detection approach most suitable for a given application will depend on the amount of anomalous data available, and whether you can distinguish anomalies from normal data. The first step in anomaly detection is to examine the data you have. Consider the following questions:<\/p>\n<h3><\/h3>\n<h3><span style=\"color: #e67e23;\"><strong>Can you see anomalies in the raw data?<\/strong><\/span><\/h3>\n<p>Sometimes you can perform anomaly detection just by looking at your data. For example, the signals in Figure 1 below were collected from a fan, and you can easily see the abrupt signal changes that indicate anomalies in the fan behavior. If you are able to detect anomalies by eye, you may be able to use a simple algorithm such as<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/help\/signal\/ref\/findchangepts.html\">findchangepts<\/a><span>\u00a0<\/span>or<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/help\/stats\/controlchart.html\">controlchart<\/a><span>\u00a0<\/span>for anomaly detection.<\/p>\n<\/div>\n<\/div>\n<div class=\"cqColumns containsResourceName section resourceClass-columns\">\n<div class=\"row\">\n<div class=\"col-xs-9\">\n<div class=\"3f675f85-2950-487b-8c05-b4f01f5475eb parsys containsResourceName resourceClass-parsys\">\n<div class=\"cqImage containsResourceName section resourceClass-image\">\n<div class=\"clearfix mw-image thumbnail\">\n<p><img decoding=\"async\" src=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.full.medium.jpg\/1653027867882.jpg\" alt=\"A MATLAB plot of data from a cooling fan, showing anomalies that are easy to spot.\" width=\"529\" height=\"328\" sizes=\"auto, (min-width: 1200px) 1140px, (min-width: 992px) 940px, calc(100vw - 30px)\" loading=\"lazy\" class=\"responsiveImage fluid_image remove_border\" srcset=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.150.medium.jpg\/1653027867882.jpg 150w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.320.medium.jpg\/1653027867882.jpg 320w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.480.medium.jpg\/1653027867882.jpg 480w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.620.medium.jpg\/1653027867882.jpg 620w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.full.medium.jpg\/1653027867882.jpg 811w\" style=\"display: block; margin-left: auto; margin-right: auto;\" \/><\/p>\n<div class=\"caption\" style=\"padding-left: 80px;\"><em><strong>Figure 1: <\/strong>A MATLAB plot of data from a cooling fan, showing anomalies that are easy to spot.<\/em><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\">\n<h3><\/h3>\n<h3><strong><span style=\"color: #e67e23;\">Can you see anomalies in derived features?<\/span><\/strong><\/h3>\n<p>Anomalies are often difficult to detect visually from raw data. In the signals below, it is difficult to determine which of the time domain signals is anomalous. However, if you create a power spectrum to view the data in the frequency domain, differences in the frequency and magnitude of the peaks clearly show that the two signals are quite different. In such cases, you can use these peaks as feature inputs to an anomaly detection algorithm based on<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/discovery\/supervised-learning.html\">supervised learning<\/a><span>\u00a0<\/span>methods.<\/p>\n<\/div>\n<div class=\"mw-text\"><\/div>\n<\/div>\n<div class=\"cqColumns containsResourceName section resourceClass-columns\">\n<div class=\"row\">\n<div class=\"col-xs-9\">\n<div class=\"dc270f83-582f-4e45-8f03-d647cc3874a6 parsys containsResourceName resourceClass-parsys\">\n<div class=\"cqImage containsResourceName section resourceClass-image\">\n<div class=\"clearfix mw-image thumbnail\" style=\"padding-left: 80px;\">\n<p><img decoding=\"async\" src=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_9539924\/dc270f83-582f-4e45-8f03-d647cc3874a6\/image_copy.adapt.full.medium.jpg\/1653027868054.jpg\" alt=\"Although anomalies are not visually apparent in the raw time series signals (eft), viewing the data in the frequency domain (right, using a periodogram in MATLAB) shows clear differences in the peak frequencies.\" width=\"694\" height=\"482\" sizes=\"auto, (min-width: 1200px) 1140px, (min-width: 992px) 940px, calc(100vw - 30px)\" loading=\"lazy\" class=\"responsiveImage remove_border\" srcset=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_9539924\/dc270f83-582f-4e45-8f03-d647cc3874a6\/image_copy.adapt.150.medium.jpg\/1653027868054.jpg 150w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_9539924\/dc270f83-582f-4e45-8f03-d647cc3874a6\/image_copy.adapt.320.medium.jpg\/1653027868054.jpg 320w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_9539924\/dc270f83-582f-4e45-8f03-d647cc3874a6\/image_copy.adapt.480.medium.jpg\/1653027868054.jpg 480w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_9539924\/dc270f83-582f-4e45-8f03-d647cc3874a6\/image_copy.adapt.620.medium.jpg\/1653027868054.jpg 620w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_9539924\/dc270f83-582f-4e45-8f03-d647cc3874a6\/image_copy.adapt.full.medium.jpg\/1653027868054.jpg 881w\" \/><\/p>\n<div class=\"caption\"><em><strong>Figure 2:<\/strong> Although anomalies are not visually apparent in the raw time series signals (left), viewing the data in the frequency domain (right, using a periodogram in MATLAB) shows clear differences in the peak frequencies.<\/em><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\">\n<h3><\/h3>\n<h3><span style=\"color: #e67e23;\"><strong>Can you statistically separate normal and anomalous features?<\/strong><\/span><\/h3>\n<p>Anomalies are not always apparent in a single signal. Today\u2019s complex machines can have hundreds of sensors, and sometimes anomalies become apparent only when considering multiple sensors at once. When you have labeled data, you can examine statistical distributions of time- and frequency-domain features, as shown in Figure 3. You may also perform<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/help\/stats\/dimensionality-reduction.html\">feature transformation<\/a><span>\u00a0<\/span>and ranking to identify the features that best separate the two groups. Then, you can use these features to train an anomaly detection algorithm on the labeled data using supervised learning.<\/p>\n<\/div>\n<div class=\"mw-text\" style=\"padding-left: 40px;\"><\/div>\n<\/div>\n<div class=\"cqImage containsResourceName section resourceClass-image\" style=\"padding-left: 40px;\">\n<div class=\"clearfix mw-image thumbnail\" style=\"padding-left: 40px;\">\n<p><strong><img decoding=\"async\" src=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/image_copy_copy.adapt.full.medium.jpg\/1653027868113.jpg\" alt=\"On the left, the MATLAB plots represent paired normal and anomalous data, in blue and red, respectively. At right, corresponding feature histograms from the Diagnostic Feature Designer are used to identify which features clearly separate normal and anomalous data for supervised anomaly detection algorithms.\" width=\"1073\" height=\"388\" sizes=\"auto, (min-width: 1200px) 1140px, (min-width: 992px) 940px, calc(100vw - 30px)\" loading=\"lazy\" class=\"responsiveImage remove_border\" srcset=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/image_copy_copy.adapt.150.medium.jpg\/1653027868113.jpg 150w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/image_copy_copy.adapt.320.medium.jpg\/1653027868113.jpg 320w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/image_copy_copy.adapt.480.medium.jpg\/1653027868113.jpg 480w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/image_copy_copy.adapt.620.medium.jpg\/1653027868113.jpg 620w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/image_copy_copy.adapt.full.medium.jpg\/1653027868113.jpg 1200w\" \/><\/strong><\/p>\n<div class=\"caption\"><em><strong>Figure 3:<\/strong> On the left, the MATLAB plots represent paired normal and anomalous data, in blue and red, respectively. On the right, corresponding feature histograms from the Diagnostic Feature Designer are used to identify which features clearly separate normal and anomalous data for supervised anomaly detection algorithms.<\/em><\/div>\n<\/div>\n<\/div>\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\">\n<h3><\/h3>\n<h3><span style=\"color: #e67e23;\"><strong>What if you don\u2019t know what an anomaly looks like?<\/strong><\/span><\/h3>\n<p>Machinery downtime is expensive, so operators often aim to prevent problems with a conservative maintenance schedule. This can mean anomalies are rare, which makes designing an anomaly detection algorithm tricky. Several approaches to designing anomaly detection algorithms require little or no anomalous data. These \u201cnormal-only\u201d methods train an algorithm on normal data only, and identify data outside those norms as anomalous. With MATLAB,\u00a0you can apply the following normal-only anomaly detection approaches to your data:<\/p>\n<ul>\n<li><b>Thresholding.<span>\u00a0<\/span><\/b>Thresholding<b>\u00a0<\/b>identifies an anomaly when data exceeds a threshold on a statistical metric. Examples include the standard deviation over recent windows in time series data, using a control chart on a signal, finding abrupt changes in a signal using change point detection, or obtaining robust estimates of the data distribution and identifying anomalies as samples on the fringes of the distribution. Thresholding on statistical metrics can be a good start, but this approach is more difficult to apply to multivariate data and less robust than machine learning approaches to anomaly detection. Statistical estimates that are robust to outliers will yield better results, such as\u00a0<a href=\"https:\/\/uk.mathworks.com\/help\/stats\/robustcov.html\">robust covariance<\/a>.<\/li>\n<li><b>One-Class Support Vector Machines.<span>\u00a0<\/span><\/b>One-class\u00a0<a href=\"https:\/\/uk.mathworks.com\/help\/stats\/fitcsvm.html\">support vector machines<\/a><span>\u00a0<\/span>identify separating hyper planes that maximize the distance between classes. Training only one class yields a model of data that can be considered normal, which allows you to detect anomalies without having any labeled anomalies available for training. This approach and other distance-based methods require numeric features as input and will not work well on high-dimensional data.<\/li>\n<li><b>Isolation Forests.<\/b>\u00a0<a href=\"https:\/\/uk.mathworks.com\/help\/stats\/iforest.html\">Isolation forests<\/a><span>\u00a0<\/span>build trees that isolate each observation into a leaf, and an anomaly score is computed as the average depth to your sample: normal samples take fewer decisions than anomalous ones. This method supports a mix of numeric and categorical features and works on high-dimensional data.<\/li>\n<li><b>Autoencoders.<\/b>\u00a0<a href=\"https:\/\/uk.mathworks.com\/discovery\/autoencoder.html\">Autoencoders<\/a>\u00a0are neural networks trained on normal data that attempt to reconstruct the original input. The trained autoencoder will reconstruct a normal input accurately. A large difference between the input and its reconstruction could indicate an anomaly. Autoencoders can be used for both signal and image data.<\/li>\n<\/ul>\n<\/div>\n<\/div>\n<div class=\"cqColumns containsResourceName section resourceClass-columns\">\n<div class=\"row\">\n<div class=\"col-xs-12\">\n<div class=\"48508e5b-6da4-42ba-936b-6c035cc2d88c parsys containsResourceName resourceClass-parsys\">\n<div class=\"cqImage containsResourceName section resourceClass-image\">\n<div class=\"clearfix mw-image thumbnail\" style=\"padding-left: 80px;\">\n<p><img decoding=\"async\" src=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_847808623\/48508e5b-6da4-42ba-936b-6c035cc2d88c\/image_copy_copy_1382.adapt.full.medium.jpg\/1653027868284.jpg\" alt=\"Autoencoders are trained to replicate inputs. The differences between the input and its reconstruction can be used for anomaly detection in signal or image data.\" width=\"720\" height=\"341\" sizes=\"auto, (min-width: 1200px) 1140px, (min-width: 992px) 940px, calc(100vw - 30px)\" loading=\"lazy\" class=\"responsiveImage remove_border\" srcset=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_847808623\/48508e5b-6da4-42ba-936b-6c035cc2d88c\/image_copy_copy_1382.adapt.150.medium.jpg\/1653027868284.jpg 150w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_847808623\/48508e5b-6da4-42ba-936b-6c035cc2d88c\/image_copy_copy_1382.adapt.320.medium.jpg\/1653027868284.jpg 320w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_847808623\/48508e5b-6da4-42ba-936b-6c035cc2d88c\/image_copy_copy_1382.adapt.480.medium.jpg\/1653027868284.jpg 480w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_847808623\/48508e5b-6da4-42ba-936b-6c035cc2d88c\/image_copy_copy_1382.adapt.620.medium.jpg\/1653027868284.jpg 620w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns_847808623\/48508e5b-6da4-42ba-936b-6c035cc2d88c\/image_copy_copy_1382.adapt.full.medium.jpg\/1653027868284.jpg 811w\" \/><\/p>\n<div class=\"caption\"><em><strong>Figure 4:<\/strong> Autoencoders are trained to replicate inputs. The differences between the input and its reconstruction can be used for anomaly detection in signal or image data.<\/em><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\"><\/div>\n<h3 class=\"mw-text\"><\/h3>\n<div class=\"mw-text\">When you have anomalies in your data but cannot label them, you can also try<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/discovery\/unsupervised-learning.html\">unsupervised clustering<\/a><span>\u00a0<\/span>approaches to anomaly detection. Sometimes you can associate clusters with normal and anomalous data, but unless your dataset is balanced (containing many anomalies of the same type), useful results are more likely with the normal-only methods. Prior to applying anomaly detection algorithms, you often need to extract features from your raw data. MATLAB supports both manual and automated methods for extracting features from signal, image, and text data. The<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/help\/predmaint\/ref\/diagnosticfeaturedesigner-app.html\">Diagnostic Feature Designer<\/a><span>\u00a0<\/span>in<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/products\/predictive-maintenance.html\">Predictive Maintenance Toolbox\u2122<\/a><span>\u00a0<\/span>can help you extract features from many types of signals.<\/div>\n<h3 class=\"mw-text\"><strong style=\"color: #e67e23; font-size: 1.75rem;\"><\/strong><\/h3>\n<h3><span style=\"color: #e67e23;\"><strong>Key Points<\/strong><\/span><strong style=\"color: #e67e23; font-size: 1.75rem;\"><\/strong><\/h3>\n<\/div>\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\">\n<ul>\n<li>Anomaly detection helps you identify outliers, deviations from normal, and unexpected behaviors<\/li>\n<li>If you have sufficient labeled data (including anomalies), use supervised learning for anomaly detection<\/li>\n<li>If you have mostly normal data, apply one of the specialized normal-only anomaly detection approaches<\/li>\n<\/ul>\n<\/div>\n<\/div>\n<div class=\"cqPanel containsResourceName resourceClass-panel section\">\n<div class=\"panel panel-default add_background_color_gray add_margin_40 section_downsize\">\n<div class=\"panel-body\">\n<div class=\"panelParsys parsys containsResourceName resourceClass-parsys\">\n<div class=\"text containsResourceName section resourceClass-text\">\n<div class=\"mw-text\">\n<p><span style=\"color: #e67e23;\"><b><\/b><\/span><\/p>\n<h3><strong><span style=\"color: #e67e23;\"><\/span><\/strong><\/h3>\n<h3><strong><span style=\"color: #e67e23;\">Detect Object Defects in Images with MATLAB<\/span><\/strong><\/h3>\n<p>To build an image-based anomaly detection algorithm, you can: 1) feed your images through the pretrained AlexNet convolutional neural network, 2) use the network activations after the first layers as features, and then 3) train a one-class support vector machine with<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/help\/stats\/fitcsvm.html\">fitcsvm<\/a>. The one-class SVM is trained on normal images, and negative classification scores indicate anomalies. In the example below, the trained model correctly identified the four hex nuts with surface defects. Try it in<span>\u00a0<\/span><a href=\"https:\/\/uk.mathworks.com\/matlabcentral\/fileexchange\/64070-deep-learning-image-anomaly-detection-for-production-line\">this example<\/a>.<\/p>\n<\/div>\n<div class=\"mw-text\"><\/div>\n<\/div>\n<div class=\"cqImage containsResourceName section resourceClass-image\">\n<div class=\"row\">\n<div class=\"col-xs-12 col-sm-6\">\n<div class=\"clearfix mw-image thumbnail\" style=\"padding-left: 40px;\"><img decoding=\"async\" src=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/panel_copy_copy\/panelParsys\/image.adapt.full.medium.jpg\/1653027868634.jpg\" alt=\"Detect object defects in images with MATLAB\" width=\"435\" height=\"435\" sizes=\"auto, (min-width: 1200px) 1140px, (min-width: 992px) 940px, calc(100vw - 30px)\" loading=\"lazy\" class=\"responsiveImage fluid_image remove_border\" srcset=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/panel_copy_copy\/panelParsys\/image.adapt.150.medium.jpg\/1653027868634.jpg 150w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/panel_copy_copy\/panelParsys\/image.adapt.320.medium.jpg\/1653027868634.jpg 320w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/panel_copy_copy\/panelParsys\/image.adapt.480.medium.jpg\/1653027868634.jpg 480w, https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/panel_copy_copy\/panelParsys\/image.adapt.full.medium.jpg\/1653027868634.jpg 613w\" \/><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<h3 style=\"text-align: justify;\"><\/h3>\n<h3><\/h3>\n<p style=\"text-align: justify;\"><a class=\"maxbutton-4 maxbutton maxbutton-download-a-free-trial\" target=\"_blank\" rel=\"noopener\" href=\"https:\/\/ascendas-asia.com\/th\/matlab-trial-for-predictive-maintenance\/\"><span class='mb-text'>Download a FREE Trial<\/span><\/a>\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0<a class=\"maxbutton-1 maxbutton maxbutton-get-quote\" target=\"_blank\" rel=\"noopener\" href=\"https:\/\/ascendas-asia.com\/th\/company\/#contact-us\"><span class='mb-text'>Request Consultation<\/span><\/a><\/p>","protected":false},"excerpt":{"rendered":"<p>Identify unexpected events and departures from normal behavior Anomaly detection is the process of identifying events or patterns that differ from expected behavior. Anomaly detection can range from simple outlier detection to complex machine learning algorithms trained to uncover hidden patterns across hundreds of signals. Engineers and data scientists use anomaly detection to identify: Faults [&hellip;]<\/p>","protected":false},"author":4,"featured_media":0,"parent":18,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"content-type":"","footnotes":"","_links_to":"","_links_to_target":""},"class_list":["post-4338","page","type-page","status-publish","hentry"],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v22.1 (Yoast SEO v27.7) - https:\/\/yoast.com\/product\/yoast-seo-premium-wordpress\/ -->\n<title>Anomaly Detection - TechSource Systems &amp; Ascendas Systems Group<\/title>\n<meta name=\"description\" content=\"Identify unexpected events and departures from normal behavior. Anomaly detection is the process of identifying events or patterns\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/ascendas-asia.com\/th\/resources\/anomaly-detection\/\" \/>\n<meta property=\"og:locale\" content=\"th_TH\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Anomaly Detection\" \/>\n<meta property=\"og:description\" content=\"Identify unexpected events and departures from normal behavior. Anomaly detection is the process of identifying events or patterns\" \/>\n<meta property=\"og:url\" content=\"https:\/\/ascendas-asia.com\/th\/resources\/anomaly-detection\/\" \/>\n<meta property=\"og:site_name\" content=\"TechSource Systems &amp; Ascendas Systems Group\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/techsourcesystems\" \/>\n<meta property=\"article:modified_time\" content=\"2022-05-27T02:40:08+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.full.medium.jpg\/1653027867882.jpg\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"6 \u0e19\u0e32\u0e17\u0e35\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/\",\"url\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/\",\"name\":\"Anomaly Detection - TechSource Systems &amp; Ascendas Systems Group\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/uk.mathworks.com\\\/discovery\\\/anomaly-detection\\\/_jcr_content\\\/mainParsys\\\/columns\\\/3f675f85-2950-487b-8c05-b4f01f5475eb\\\/image.adapt.full.medium.jpg\\\/1653027867882.jpg\",\"datePublished\":\"2022-05-24T05:42:47+00:00\",\"dateModified\":\"2022-05-27T02:40:08+00:00\",\"description\":\"Identify unexpected events and departures from normal behavior. Anomaly detection is the process of identifying events or patterns\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/#breadcrumb\"},\"inLanguage\":\"th\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"th\",\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/#primaryimage\",\"url\":\"https:\\\/\\\/uk.mathworks.com\\\/discovery\\\/anomaly-detection\\\/_jcr_content\\\/mainParsys\\\/columns\\\/3f675f85-2950-487b-8c05-b4f01f5475eb\\\/image.adapt.full.medium.jpg\\\/1653027867882.jpg\",\"contentUrl\":\"https:\\\/\\\/uk.mathworks.com\\\/discovery\\\/anomaly-detection\\\/_jcr_content\\\/mainParsys\\\/columns\\\/3f675f85-2950-487b-8c05-b4f01f5475eb\\\/image.adapt.full.medium.jpg\\\/1653027867882.jpg\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/anomaly-detection\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/ascendas-asia.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Resources\",\"item\":\"https:\\\/\\\/ascendas-asia.com\\\/resources\\\/\"},{\"@type\":\"ListItem\",\"position\":3,\"name\":\"Anomaly Detection\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/#website\",\"url\":\"https:\\\/\\\/ascendas-asia.com\\\/\",\"name\":\"TechSource Systems & Ascendas Systems Group | MathWorks Authorized Reseller\",\"description\":\"\",\"publisher\":{\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/ascendas-asia.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"th\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/#organization\",\"name\":\"TechSource Systems & Ascendas Systems Group\",\"url\":\"https:\\\/\\\/ascendas-asia.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"th\",\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/ascendas-asia.com\\\/wp-content\\\/uploads\\\/2021\\\/12\\\/logo.jpg\",\"contentUrl\":\"https:\\\/\\\/ascendas-asia.com\\\/wp-content\\\/uploads\\\/2021\\\/12\\\/logo.jpg\",\"width\":825,\"height\":131,\"caption\":\"TechSource Systems & Ascendas Systems Group\"},\"image\":{\"@id\":\"https:\\\/\\\/ascendas-asia.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/www.facebook.com\\\/techsourcesystems\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/techsource-systems\\\/\",\"https:\\\/\\\/www.youtube.com\\\/c\\\/TechSourceSystems\"]}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Anomaly Detection - TechSource Systems &amp; Ascendas Systems Group","description":"Identify unexpected events and departures from normal behavior. Anomaly detection is the process of identifying events or patterns","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/ascendas-asia.com\/th\/resources\/anomaly-detection\/","og_locale":"th_TH","og_type":"article","og_title":"Anomaly Detection","og_description":"Identify unexpected events and departures from normal behavior. Anomaly detection is the process of identifying events or patterns","og_url":"https:\/\/ascendas-asia.com\/th\/resources\/anomaly-detection\/","og_site_name":"TechSource Systems &amp; Ascendas Systems Group","article_publisher":"https:\/\/www.facebook.com\/techsourcesystems","article_modified_time":"2022-05-27T02:40:08+00:00","og_image":[{"url":"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.full.medium.jpg\/1653027867882.jpg","type":"","width":"","height":""}],"twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"6 \u0e19\u0e32\u0e17\u0e35"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/","url":"https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/","name":"Anomaly Detection - TechSource Systems &amp; Ascendas Systems Group","isPartOf":{"@id":"https:\/\/ascendas-asia.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/#primaryimage"},"image":{"@id":"https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/#primaryimage"},"thumbnailUrl":"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.full.medium.jpg\/1653027867882.jpg","datePublished":"2022-05-24T05:42:47+00:00","dateModified":"2022-05-27T02:40:08+00:00","description":"Identify unexpected events and departures from normal behavior. Anomaly detection is the process of identifying events or patterns","breadcrumb":{"@id":"https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/#breadcrumb"},"inLanguage":"th","potentialAction":[{"@type":"ReadAction","target":["https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/"]}]},{"@type":"ImageObject","inLanguage":"th","@id":"https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/#primaryimage","url":"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.full.medium.jpg\/1653027867882.jpg","contentUrl":"https:\/\/uk.mathworks.com\/discovery\/anomaly-detection\/_jcr_content\/mainParsys\/columns\/3f675f85-2950-487b-8c05-b4f01f5475eb\/image.adapt.full.medium.jpg\/1653027867882.jpg"},{"@type":"BreadcrumbList","@id":"https:\/\/ascendas-asia.com\/resources\/anomaly-detection\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/ascendas-asia.com\/"},{"@type":"ListItem","position":2,"name":"Resources","item":"https:\/\/ascendas-asia.com\/resources\/"},{"@type":"ListItem","position":3,"name":"Anomaly Detection"}]},{"@type":"WebSite","@id":"https:\/\/ascendas-asia.com\/#website","url":"https:\/\/ascendas-asia.com\/","name":"TechSource Systems & Ascendas Systems Group | MathWorks Authorized Reseller","description":"","publisher":{"@id":"https:\/\/ascendas-asia.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/ascendas-asia.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"th"},{"@type":"Organization","@id":"https:\/\/ascendas-asia.com\/#organization","name":"TechSource Systems & Ascendas Systems Group","url":"https:\/\/ascendas-asia.com\/","logo":{"@type":"ImageObject","inLanguage":"th","@id":"https:\/\/ascendas-asia.com\/#\/schema\/logo\/image\/","url":"https:\/\/ascendas-asia.com\/wp-content\/uploads\/2021\/12\/logo.jpg","contentUrl":"https:\/\/ascendas-asia.com\/wp-content\/uploads\/2021\/12\/logo.jpg","width":825,"height":131,"caption":"TechSource Systems & Ascendas Systems Group"},"image":{"@id":"https:\/\/ascendas-asia.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/techsourcesystems","https:\/\/www.linkedin.com\/company\/techsource-systems\/","https:\/\/www.youtube.com\/c\/TechSourceSystems"]}]}},"_links":{"self":[{"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/pages\/4338","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/comments?post=4338"}],"version-history":[{"count":14,"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/pages\/4338\/revisions"}],"predecessor-version":[{"id":4514,"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/pages\/4338\/revisions\/4514"}],"up":[{"embeddable":true,"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/pages\/18"}],"wp:attachment":[{"href":"https:\/\/ascendas-asia.com\/th\/wp-json\/wp\/v2\/media?parent=4338"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}