{"id":1239,"date":"2017-01-13T07:26:29","date_gmt":"2017-01-13T07:26:29","guid":{"rendered":"http:\/\/dataaspirant.com\/?p=1239"},"modified":"2017-02-19T07:27:42","modified_gmt":"2017-02-19T07:27:42","slug":"support-vector-machine-algorithm","status":"publish","type":"post","link":"https:\/\/dataaspirant.com\/support-vector-machine-algorithm\/","title":{"rendered":"Svm classifier, Introduction to support vector machine algorithm"},"content":{"rendered":"<div id=\"bsf_rt_marker\"><\/div><div id=\"attachment_1370\" style=\"width: 1034px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1370\" class=\"size-full wp-image-1370 lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm.jpg\" alt=\"Introduction to svm (Support Vector Machine)\" width=\"1024\" height=\"768\" data-srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm.jpg 1024w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm-300x225.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm-768x576.jpg 768w\" data-sizes=\"(max-width: 1024px) 100vw, 1024px\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 1024px; --smush-placeholder-aspect-ratio: 1024\/768;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1370\" class=\"size-full wp-image-1370\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm.jpg\" alt=\"Introduction to svm (Support Vector Machine)\" width=\"1024\" height=\"768\" srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm.jpg 1024w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm-300x225.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/Support-vector-machine-svm-768x576.jpg 768w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/noscript><p id=\"caption-attachment-1370\" class=\"wp-caption-text\"><strong>vmSupport vector machine introduction<\/strong><\/p><\/div>\n<h1 style=\"text-align: center;\">SVM Classifier Introduction<\/h1>\n<p style=\"text-align: justify;\">Hi, welcome to the another post on classification concepts. So far we have talked bout different classification concepts like logistic regression, knn classifier, decision trees .., etc. In this article, we were going to discuss \u00a0support vector machine which is a <a href=\"https:\/\/dataaspirant.com\/2014\/09\/19\/supervised-and-unsupervised-learning\/\" target=\"_blank\">supervised learning algorithm<\/a>. Just to give why we were so interested to write about Svm as it is one of the powerful technique for Classification, <a href=\"https:\/\/dataaspirant.com\/2014\/10\/02\/linear-regression\/\" target=\"_blank\">Regression<\/a> \u00a0&amp; Outlier detection with an intuitive model.<\/p>\n<p style=\"text-align: justify;\">Before we drive into the concepts of support vector machine, let&#8217;s remember the backend heads of Svm classifier. <strong>Vapnik &amp; Chervonenkis<\/strong> originally invented support vector machine.\u00a0At that time, the algorithm was in early stages. Drawing hyperplanes only for linear classifier was possible.<\/p>\n<p style=\"text-align: justify;\"><!--more--><\/p>\n<p style=\"text-align: justify;\">Later in 1992 <strong>Vapnik, Boser &amp; Guyon<\/strong> suggested a way\u00a0for building a non-linear classifier. They suggested using kernel trick in SVM latest paper. Vapnik &amp; Cortes published this paper in the year 1995<\/p>\n<p style=\"text-align: justify;\">From then, Svm classifier treated as one of the dominant classification algorithms. In further sections of our article, we were going to discuss linear and non-linear classes. However, Svm is a supervised learning technique. When we have a dataset with features &amp; class labels both then we can use Support Vector Machine. But if in our dataset do not have class labels or outputs of our feature set then it is considered as an <strong>unsupervised learning algorithm<\/strong>. In that case, we can use Support Vector Clustering.<\/p>\n<p style=\"text-align: justify;\">Enough of the\u00a0<span style=\"background-color: #f5f6f5;\">introduction<\/span>\u00a0to support vector machine algorithm. Let&#8217;s drive into the key concepts.<\/p>\n<h3>How Svm classifier Works?<\/h3>\n<p style=\"text-align: justify;\">For a dataset consisting of features set and labels set, an SVM classifier builds a model to predict classes for new examples. It assigns new example\/data points to one of the classes. If there are only 2 classes then it can be called as a Binary SVM Classifier.<\/p>\n<p style=\"text-align: justify;\">There are 2 kinds of SVM classifiers:<\/p>\n<ol>\n<li><strong>Linear SVM Classifier<\/strong><\/li>\n<li><strong>Non-Linear SVM Classifier<\/strong><\/li>\n<\/ol>\n<h4 style=\"text-align: justify;\">Svm Linear Classifier:<\/h4>\n<p style=\"text-align: justify;\">In the\u00a0linear classifier\u00a0model, we assumed that training examples plotted in space. These data points are expected to be separated by an apparent gap. It predicts a straight hyperplane dividing 2 classes. The primary focus while drawing the hyperplane is on maximizing the distance from hyperplane to the nearest data point of either class. The drawn hyperplane called as a\u00a0maximum-margin hyperplane.<\/p>\n<h4 style=\"text-align: justify;\">SVM Non-Linear\u00a0Classifier:<\/h4>\n<p style=\"text-align: justify;\">\u00a0In the\u00a0real world, our dataset is generally dispersed up to some extent. To solve this problem separation of data into different classes on the basis of a\u00a0straight linear hyperplane can&#8217;t be considered a good choice. For this Vapnik suggested creating Non-Linear Classifiers by applying the kernel trick to maximum-margin hyperplanes. In Non-Linear SVM Classification, data points plotted in a higher dimensional space.<\/p>\n<h3>Examples of SVM boundaries<\/h3>\n<p style=\"text-align: justify;\">In this section, we will learn about selecting best hyperplane for our classification. We will show data from 2 classes. The classes represented by triangle $latex \\bigtriangleup$ and circle$latex \\bigcirc$.<\/p>\n<table style=\"height: 163px; background-color: #ffffff;\" width=\"796\">\n<tbody>\n<tr style=\"background-color: #ffffff;\">\n<td style=\"width: 390px;\">\n<h5>1.\u00a0Case 1:<\/h5>\n<p style=\"text-align: justify;\">Consider the case in Fig 2, with data from 2 different classes.Now, we wish to find the best hyperplane which can separate the two classes.<\/p>\n<p style=\"text-align: justify;\">Please check Fig 1. on the right to find which hyperplane best suit this use case.In SVM, we try to maximize the distance between hyperplane &amp; nearest data point. This is known as margin.<\/p>\n<p style=\"text-align: justify;\">Since 1st decision boundary is maximizing the distance between classes on left and right. So, our maximum margin hyperplane will be &#8220;1st &#8220;.<\/p>\n<\/td>\n<td style=\"width: 390px; text-align: center;\">\n<p><div id=\"attachment_1417\" style=\"width: 310px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1417\" class=\"wp-image-1417 size-medium lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-300x300.jpg\" width=\"300\" height=\"300\" data-srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1.jpg 800w\" data-sizes=\"(max-width: 300px) 100vw, 300px\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 300px; --smush-placeholder-aspect-ratio: 300\/300;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1417\" class=\"wp-image-1417 size-medium\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-300x300.jpg\" width=\"300\" height=\"300\" srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-1.jpg 800w\" sizes=\"(max-width: 300px) 100vw, 300px\" \/><\/noscript><p id=\"caption-attachment-1417\" class=\"wp-caption-text\">Fig 1<\/p><\/div><\/td>\n<\/tr>\n<tr style=\"background-color: #ffffff;\">\n<td style=\"width: 390px;\">\n<h5>Case 2:<\/h5>\n<p style=\"text-align: justify;\">Consider the case in Fig 2, with data from 2 different classes.Now, we wish to\u00a0find the best hyperplane which can separate the two classes.<\/p>\n<p style=\"text-align: justify;\">As data of each class is distributed either on left\u00a0or right. Our motive is to select hyperplane which can separate the classes with maximum margin.<\/p>\n<p style=\"text-align: justify;\">In this case, all the decision boundaries are separating classes but\u00a0only <strong>1st<\/strong> decision boundary is showing maximum margin between $latex \\bigtriangleup$ &amp; $latex \\bigcirc $.<\/p>\n<\/td>\n<td style=\"width: 390px; text-align: center;\">\n<p><div id=\"attachment_1419\" style=\"width: 310px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1419\" class=\"wp-image-1419 size-medium lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-300x300.jpg\" width=\"300\" height=\"300\" data-srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3.jpg 800w\" data-sizes=\"(max-width: 300px) 100vw, 300px\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 300px; --smush-placeholder-aspect-ratio: 300\/300;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1419\" class=\"wp-image-1419 size-medium\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-300x300.jpg\" width=\"300\" height=\"300\" srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-3.jpg 800w\" sizes=\"(max-width: 300px) 100vw, 300px\" \/><\/noscript><p id=\"caption-attachment-1419\" class=\"wp-caption-text\">Fig 2<\/p><\/div><\/td>\n<\/tr>\n<tr style=\"background-color: #ffffff;\">\n<td style=\"width: 390px;\">\n<h5>Case 3:<\/h5>\n<p style=\"text-align: justify;\">Consider the case in Fig 3, with data from 2<br \/>\ndifferent\u00a0classes.\u00a0Now, we wish to find the best hyperplane which can separate the two classes.<\/p>\n<p style=\"text-align: justify;\">Data is not evenly distributed on left and right. Some of the $latex \\bigtriangleup $ are on right too. You may feel we can ignore the two data points above 3rd hyperplane but\u00a0that would be incorrect.<\/p>\n<p style=\"text-align: justify;\">SVM tries to find out maximum margin hyperplane but gives first priority to\u00a0correct classification.<br \/>\n1st decision boundary is separating some\u00a0$latex \\bigtriangleup $ from $latex \\bigcirc $ but not all. It&#8217;s not even showing good margin.<\/p>\n<p style=\"text-align: justify;\">2nd decision boundary is separating the data points similar to 1st boundary but here margin between boundary and data points is larger than the\u00a0previous case.<\/p>\n<p style=\"text-align: justify;\">3rd decision boundary is separating all $latex \\bigtriangleup $ from all $latex \\bigcirc $ classes.<br \/>\nSo,\u00a0SVM will select <strong>3rd<\/strong>\u00a0hyperplane.<\/p>\n<\/td>\n<td style=\"width: 390px; text-align: center;\">\n<p><div id=\"attachment_1420\" style=\"width: 310px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1420\" class=\"wp-image-1420 size-medium lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-300x300.jpg\" width=\"300\" height=\"300\" data-srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4.jpg 800w\" data-sizes=\"(max-width: 300px) 100vw, 300px\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 300px; --smush-placeholder-aspect-ratio: 300\/300;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1420\" class=\"wp-image-1420 size-medium\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-300x300.jpg\" width=\"300\" height=\"300\" srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-4.jpg 800w\" sizes=\"(max-width: 300px) 100vw, 300px\" \/><\/noscript><p id=\"caption-attachment-1420\" class=\"wp-caption-text\">Figure 3<\/p><\/div><\/td>\n<\/tr>\n<tr style=\"background-color: #ffffff;\">\n<td style=\"width: 390px;\">\n<h5>Case 4:<\/h5>\n<p style=\"text-align: justify;\">Consider the figure 4, we will learn about outliers in SVM.<\/p>\n<p style=\"text-align: justify;\">We wish to find the best hyperplane which can separate the two classes.<br \/>\nData is not evenly distributed on left and right. Some of the $latex \\bigtriangleup $ are on right too.<\/p>\n<p style=\"text-align: justify;\">In the\u00a0real world, you may find f ew values that correspond\u00a0to extreme cases\u00a0i.e, exceptions.These exceptions are known as Outliers. SVM have the capability to detect and ignore\u00a0outliers.In the image, 2 $latex \\bigtriangleup \\text{&#8216;s}$ are in between the group of $latex \\bigcirc $. These\u00a0$latex \\bigtriangleup \\text{&#8216;s}$ are outliers.<\/p>\n<p style=\"text-align: justify;\">While selecting hyperplane, SVM will automatically ignore these\u00a0$latex \\bigtriangleup \\text{&#8216;s}$ and select best-performing hyperplane.1st &amp; 2nd decision boundaries are separating classes but 1st decision boundary shows maximum margin in between boundary and support vectors.<\/p>\n<\/td>\n<td style=\"width: 390px; text-align: center;\">\n<p><div id=\"attachment_1421\" style=\"width: 310px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1421\" class=\"wp-image-1421 size-medium lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-300x300.jpg\" width=\"300\" height=\"300\" data-srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5.jpg 800w\" data-sizes=\"(max-width: 300px) 100vw, 300px\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 300px; --smush-placeholder-aspect-ratio: 300\/300;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1421\" class=\"wp-image-1421 size-medium\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-300x300.jpg\" width=\"300\" height=\"300\" srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-300x300.jpg 300w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-150x150.jpg 150w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5-768x768.jpg 768w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-5.jpg 800w\" sizes=\"(max-width: 300px) 100vw, 300px\" \/><\/noscript><p id=\"caption-attachment-1421\" class=\"wp-caption-text\">Figure 4<\/p><\/div><\/td>\n<\/tr>\n<tr style=\"background-color: #ffffff;\">\n<td style=\"width: 390px;\">\n<h5>Case 5:<strong>\u00a0<\/strong><\/h5>\n<p style=\"text-align: justify;\">We will learn about non-linear classifiers.<\/p>\n<p style=\"text-align: justify;\">Please check the figure 5 on right. It&#8217;s showing that data can&#8217;t be<br \/>\nseparated by any straight line, i.e, data is not linearly separable.SVM possess the option of using Non-Linear classifier.<\/p>\n<p style=\"text-align: justify;\">We can use different types of kernels like Radial Basis Function Kernel, Polynomial kernel etc. We have shown a decision boundary separating both the classes. This decision boundary resembles a parabola.<\/p>\n<\/td>\n<td style=\"width: 390px; text-align: center;\">\n<p><div style=\"width: 310px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-6-300x300.jpg\" width=\"300\" height=\"300\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" class=\"lazyload\" style=\"--smush-placeholder-width: 300px; --smush-placeholder-aspect-ratio: 300\/300;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/SVM-Intro-6-300x300.jpg\" width=\"300\" height=\"300\" \/><\/noscript><p class=\"wp-caption-text\">Figure 5<\/p><\/div><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n<h3><strong>Linear Support Vector Machine Classifier<\/strong><\/h3>\n<p style=\"text-align: justify;\">In Linear Classifier, A data point considered as a p-dimensional vector(list of p-numbers) and we separate points using (p-1) dimensional hyperplane. There can be many hyperplanes separating data in a linear order, but the best hyperplane is considered to be the one which maximizes the margin i.e., the distance between hyperplane and closest data point of either class.<\/p>\n<p style=\"text-align: justify;\">The\u00a0Maximum-margin hyperplane is determined by the data points that lie nearest to it. Since we have to maximize the distance between hyperplane and the data points. These data points which influences our hyperplane are known as support vectors.<\/p>\n<h3>Non-Linear\u00a0Support Vector Machine Classifier<\/h3>\n<p style=\"text-align: justify;\">Vapnik proposed Non-Linear Classifiers in 1992. \u00a0It often happens that our data points are not linearly separable in a p-dimensional(finite) space. To solve this, it was proposed to map p-dimensional space into a much higher dimensional space. We can draw customized\/non-linear hyperplanes using Kernel trick.<br \/>\nEvery kernel holds a non-linear kernel function.<\/p>\n<p style=\"text-align: justify;\">This function helps to build a high dimensional feature space. There are many kernels that have been developed. Some standard kernels are:<\/p>\n<ol>\n<li><strong>Polynomial (homogeneous) Kernel:<\/strong><img decoding=\"async\" class=\"mwe-math-fallback-image-inline aligncenter\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/d08d7e9325069d9baab53bdde0ece3da72d750ad\" alt=\"{\\displaystyle k({\\vec {x_{i}}},{\\vec {x_{j}}})=({\\vec {x_{i}}}\\cdot {\\vec {x_{j}}})^{d}}\" \/>The polynomial kernel function can be represented by the above expression. Where\u00a0k(x<sub>i<\/sub>, x<sub>j<\/sub>) is a kernel function, x<sub>i\u00a0<\/sub>&amp; x<sub>j<\/sub>\u00a0 are vectors of feature space and d is the degree of polynomial function.<\/li>\n<li><strong>Polynomial(non-homogeneous) Kernel:<\/strong><br \/>\n<img loading=\"lazy\" decoding=\"async\" class=\"mwe-math-fallback-image-inline aligncenter\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/c7bc5dc859db1eb2debb5e1521c954ca2829c9a7\" alt=\"K(x,y) = (x^\\mathsf{T} y + c)^{d}\" width=\"163\" height=\"24\" \/>In the\u00a0non-homogeneous kernel, a constant term is also added. The constant term &#8220;c&#8221; is also known as a free parameter. It influences the combination of features. x &amp; y are vectors of feature space.<\/li>\n<li><strong>Radial Basis Function Kernel:<\/strong><br \/>\nIt is also known as RBF kernel. It is one of the most popular kernels. For distance metric squared euclidean distance is used here. It is used to draw completely non-linear hyperplanes.<br \/>\n<img loading=\"lazy\" decoding=\"async\" class=\"mwe-math-fallback-image-inline aligncenter\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/72e3c5b92a3d0a07b20f123a93ed63900c1afb8d\" alt=\"K(\\mathbf {x} ,\\mathbf {x'} )=\\exp \\left(-{\\frac {||\\mathbf {x} -\\mathbf {x'} ||^{2}}{2\\sigma ^{2}}}\\right)\" width=\"190\" height=\"45\" \/><br \/>\nwhere x\u00a0&amp; x&#8217; are vectors of feature space.\u00a0<img decoding=\"async\" class=\"mwe-math-fallback-image-inline\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/59f59b7c3e6fdb1d0365a494b81fb9a696138c36\" alt=\"\\sigma \" \/> is a free parameter. Selection of parameters is a critical choice. Using a typical value of the parameter can lead to overfitting our data.<\/li>\n<\/ol>\n<h3>Selecting the Svm Hyperplanes<\/h3>\n<div style=\"width: 336px\" class=\"wp-caption alignright\"><img loading=\"lazy\" decoding=\"async\" class=\"thumbimage lazyload\" data-src=\"https:\/\/upload.wikimedia.org\/wikipedia\/commons\/thumb\/2\/2a\/Svm_max_sep_hyperplane_with_margin.png\/220px-Svm_max_sep_hyperplane_with_margin.png\" width=\"326\" height=\"351\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 326px; --smush-placeholder-aspect-ratio: 326\/351;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" class=\"thumbimage\" src=\"https:\/\/upload.wikimedia.org\/wikipedia\/commons\/thumb\/2\/2a\/Svm_max_sep_hyperplane_with_margin.png\/220px-Svm_max_sep_hyperplane_with_margin.png\" width=\"326\" height=\"351\" \/><\/noscript><p class=\"wp-caption-text\"><span style=\"font-size: 8pt;\">Source: Wikipedia Support Vector Machine<\/span><\/p><\/div>\n<p style=\"text-align: justify;\"><strong>Linearly Separable:<\/strong> For the data which can be separated linearly, we select two parallel hyperplanes that separate the two classes of data, \u00a0so that distance between both the lines is maximum.\u00a0The region b\/w these two hyperplanes is known as &#8220;margin&#8221; &amp; maximum margin hyperplane is the one that lies in the middle of them.<\/p>\n<p style=\"text-align: justify;\">\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0<img loading=\"lazy\" decoding=\"async\" class=\"wp-image-1404 alignnone lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/latex2.jpg\" alt=\"latex2\" width=\"284\" height=\"73\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 284px; --smush-placeholder-aspect-ratio: 284\/73;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-1404 alignnone\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/latex2.jpg\" alt=\"latex2\" width=\"284\" height=\"73\" \/><\/noscript><\/p>\n<p>where\u00a0<img decoding=\"async\" class=\"mwe-math-fallback-image-inline\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/5b6f27a892f3053ef0bfe273f88f18351a1a18ac\" alt=\"\\|{\\vec {w}}\\|\" \/>\u00a0is normal vector to the hyperplane, \u03b8i denotes classes &amp; xi denotes features. The\u00a0Distance between two hyperplanes is <img decoding=\"async\" class=\"mwe-math-fallback-image-inline\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/ac553c94aec996dffa3369adcf285319178a474f\" alt=\"{\\tfrac {2}{\\|{\\vec {w}}\\|}}\" \/>, to maximize this distance denominator value should be minimized i.e,\u00a0<img decoding=\"async\" class=\"mwe-math-fallback-image-inline\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/5b6f27a892f3053ef0bfe273f88f18351a1a18ac\" alt=\"\\|{\\vec {w}}\\|\" \/>\u00a0should be minimized.<br \/>\nFor proper classification, we can build a combined equation:<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-1403 alignnone lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/latex1.jpg\" alt=\"latex1\" width=\"394\" height=\"43\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 394px; --smush-placeholder-aspect-ratio: 394\/43;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-1403 alignnone\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/01\/latex1.jpg\" alt=\"latex1\" width=\"394\" height=\"43\" \/><\/noscript><\/p>\n<p>Non-Linearly Separable: To build classifier for non-linear data, we try to minimize<br \/>\n<img loading=\"lazy\" decoding=\"async\" class=\"mwe-math-fallback-image-inline aligncenter\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/53b729df53f32c7fbf933b1b034a8e368037d9b5\" alt=\"{\\displaystyle \\left[{\\frac {1}{n}}\\sum _{i=1}^{n}\\max \\left(0,1-y_{i}({\\vec {w}}\\cdot {\\vec {x}}_{i}-b)\\right)\\right]+\\lambda \\lVert {\\vec {w}}\\rVert ^{2},}\" width=\"270\" height=\"45\" \/><\/p>\n<p style=\"text-align: justify;\">Here, max() method will be zero( 0 ), if x<sub>i<\/sub> is on the\u00a0correct side of the margin. For data that is on opposite side of the margin, the function&#8217;s value is proportional to the distance from the margin.<\/p>\n<p>where,\u00a0<img decoding=\"async\" class=\"mwe-math-fallback-image-inline\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/b43d0ea3c9c025af1be9128e62a18fa74bedda2a\" alt=\"\\lambda \" \/>\u00a0determines tradeoff b\/w increasing the margin size and that\u00a0<img decoding=\"async\" class=\"mwe-math-fallback-image-inline\" src=\"https:\/\/wikimedia.org\/api\/rest_v1\/media\/math\/render\/svg\/97dc64e456d28b08c449ec343111cc5c3ce39f72\" alt=\"{\\vec {x}}_{i}\" \/>\u00a0is on correct side of the\u00a0margin.<\/p>\n<h3><strong>Advantages of SVM Classifier:<\/strong><\/h3>\n<ul>\n<li>SVMs are effective when the\u00a0number of features is quite large.<\/li>\n<li>It works effectively even if the number of features are greater than the number of samples.<\/li>\n<li>Non-Linear data can also be classified using customized hyperplanes built by using kernel trick.<\/li>\n<li>It is a robust model to solve prediction problems since it maximizes margin.<\/li>\n<\/ul>\n<h3><strong>Disadvantages of SVM Classifier:<\/strong><\/h3>\n<ul>\n<li>The biggest limitation of Support Vector Machine is the choice of the kernel. The wrong choice of the kernel can lead to an increase in error percentage.<\/li>\n<li>With a greater number of samples, it starts giving poor performances.<\/li>\n<li>SVMs have good generalization performance but they can be extremely slow in the test phase.<\/li>\n<li>SVMs have high algorithmic complexity and extensive memory requirements due to the use of quadratic programming.<\/li>\n<\/ul>\n<h3>Support Vector Machine Libraries \/ Packages:<\/h3>\n<p style=\"text-align: justify;\">For implementing support vector machine on a dataset, we can use libraries. There are many libraries or packages available that can help us to implement SVM smoothly. We just need to call functions with parameters according to our need.<\/p>\n<p style=\"text-align: justify;\">In Python, we can use libraries like sklearn. For classification, Sklearn provides functions like SVC, NuSVC &amp; LinearSVC.<\/p>\n<p style=\"text-align: justify;\"><strong>SVC()<\/strong> and <strong>NuSVC()<\/strong> methods are almost similar but with some difference in parameters. We pass values of kernel parameter, gamma and C parameter etc. By default kernel parameter uses &#8220;rbf&#8221; as its value but we can pass values like &#8220;poly&#8221;, &#8220;linear&#8221;, &#8220;sigmoid&#8221; or callable function.<\/p>\n<p style=\"text-align: justify;\"><strong>LinearSVC()<\/strong> is an SVC for Classification that uses only linear kernel. In <strong>LinearSVC()<\/strong>, we don&#8217;t pass value of kernel, since it&#8217;s specifically for linear classification.<\/p>\n<p style=\"text-align: justify;\">In R programming language, we can use packages like &#8220;e1071&#8221; or &#8220;caret&#8221;. For using a package, we need to install it first. For installing &#8220;e1071&#8221;, we can type\u00a0<span class=\"lang:default decode:true crayon-inline \">install.packages(&#8220;e1071&#8221;)<\/span>\u00a0in console.<br \/>\ne1071 provides an <strong>SVM()<\/strong> method, it can be used for both regression and classification. <strong>SVM()<\/strong> method accepts data, gamma values and kernel etc.<\/p>\n<h3>SVM Applications:<\/h3>\n<p style=\"text-align: justify;\">SVMS are a byproduct of Neural Network. They are widely applied to pattern classification and regression problems. Here are some of its applications:<\/p>\n<ol>\n<li>Facial expression classification: SVMs can be used to classify facial expressions. It uses statistical models of shape and SVMs.<\/li>\n<li>Speech recognition: SVMs are used to accept keywords and reject non-keywords them and build a model to recognize speech.<\/li>\n<li>Handwritten digit recognition:\u00a0Support vector classifiers can be applied to the recognition of isolated handwritten digits optically scanned.<\/li>\n<li>Text Categorization: In information retrieval and then categorization of data using labels can be done by SVM.<\/li>\n<\/ol>\n<h4>Reference Links:<\/h4>\n<p>[1]\u00a0<a href=\"https:\/\/en.wikipedia.org\/wiki\/Support_vector_machine\" target=\"_blank\">https:\/\/en.wikipedia.org\/wiki\/Support_vector_machine<\/a><\/p>\n<p>[2]\u00a0<a href=\"http:\/\/www.support-vector-machines.org\/SVM_soft.html\" target=\"_blank\">http:\/\/www.support-vector-machines.org\/SVM_soft.html<\/a><\/p>\n<p>[3]\u00a0<a href=\"https:\/\/core.ac.uk\/download\/pdf\/6302770.pdf\" target=\"_blank\">https:\/\/core.ac.uk\/download\/pdf\/6302770.pdf<\/a><\/p>\n<p>&nbsp;<\/p>\n<h4>Follow us:<\/h4>\n<h5><strong><span style=\"color: #0000ff;\"><a style=\"color: #0000ff;\" title=\"dataaspirant\" href=\"https:\/\/www.facebook.com\/dataaspirant\" target=\"_blank\">FACEBOOK<\/a><\/span>|\u00a0<span style=\"color: #ff0000;\"><a style=\"color: #ff0000;\" title=\"dataaspirant.quora.com\" href=\"http:\/\/dataaspirant.quora.com\/\" target=\"_blank\">QUORA<\/a><\/span>\u00a0|<span style=\"color: #00ccff;\"><a style=\"color: #00ccff;\" title=\"twitter.com\/dataaspirant\" href=\"https:\/\/twitter.com\/dataaspirant\" target=\"_blank\">TWITTER<\/a><\/span>| <span style=\"color: #808000;\"><a style=\"color: #808000;\" href=\"https:\/\/plus.google.com\/104087213513583475222\" target=\"_blank\">GOOGLE+<\/a><\/span> | <span style=\"color: #000080;\"><a style=\"color: #000080;\" href=\"https:\/\/www.linkedin.com\/company\/dataaspirant\" target=\"_blank\">LINKEDIN<\/a><\/span>|\u00a0<span style=\"color: #ff9900;\"><a style=\"color: #ff9900;\" title=\"reddit dataaspirant\" href=\"http:\/\/www.reddit.com\/user\/dataaspirant\/\" target=\"_blank\">REDDIT<\/a>\u00a0<\/span>| <span style=\"color: #ff0000;\"><a style=\"color: #ff0000;\" title=\"flipboard dataaspirant\" href=\"https:\/\/flipboard.com\/@dataaspirant\/\" target=\"_blank\">FLIPBOARD<\/a><\/span> | <span style=\"color: #00ff00;\"><a style=\"color: #00ff00;\" title=\"medium dataaspirant\" href=\"https:\/\/medium.com\/@dataaspirant\" target=\"_blank\">MEDIUM<\/a>\u00a0| <a href=\"https:\/\/github.com\/saimadhu-polamuri\/DataAspirant_codes\" target=\"_blank\"><span style=\"color: #008080;\">GITHUB<\/span><\/a><\/span><\/strong><\/h5>\n<p style=\"text-align: justify;\">I hope you like this post. If you have any questions, then feel free to comment below.\u00a0 If you want me to write on one particular topic, then do tell it to me in the comments below.<\/p>\n<h3>Related Courses:<\/h3>\n<p>Do check out <a href=\"https:\/\/dataaspirant.com\/data-science-courses\/\" target=\"_blank\">unlimited data science courses<\/a><\/p>\n<p>&nbsp;<\/p>\n<table style=\"width: 737px; height: 1202px;\">\n<tbody>\n<tr>\n<td style=\"width: 269.84375px;\"><strong>Title &amp; links<\/strong><\/td>\n<td style=\"width: 203.15625px;\"><strong>Details<\/strong><\/td>\n<td style=\"width: 250px;\"><strong>What You Will Learn<\/strong><\/td>\n<\/tr>\n<tr>\n<td style=\"width: 269.84375px;\"><a href=\"http:\/\/click.linksynergy.com\/link?id=d1QCig2q3qI&amp;offerid=323058.950390&amp;type=2&amp;murl=https%3A%2F%2Fwww.udemy.com%2Fmachinelearning%2F\" target=\"_blank\">Machine Learning A-Z: Hands-On Python &amp; R In Data Science<\/a><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/ad.linksynergy.com\/fs-bin\/show?id=d1QCig2q3qI&amp;bids=323058.950390&amp;type=2&amp;subid=0\" width=\"1\" height=\"1\" border=\"0\" \/><a class=\"hasimg\" href=\"http:\/\/click.linksynergy.com\/link?id=d1QCig2q3qI&amp;offerid=323058.950390&amp;type=2&amp;murl=https%3A%2F%2Fwww.udemy.com%2Fmachinelearning%2F\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" data-src=\"https:\/\/i2.wp.com\/udemy-images.udemy.com\/course\/480x270\/950390_270f_3.jpg?w=690&amp;ssl=1\" width=\"209\" height=\"118\" border=\"0\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" class=\"lazyload\" style=\"--smush-placeholder-width: 209px; --smush-placeholder-aspect-ratio: 209\/118;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/i2.wp.com\/udemy-images.udemy.com\/course\/480x270\/950390_270f_3.jpg?w=690&amp;ssl=1\" width=\"209\" height=\"118\" border=\"0\" \/><\/noscript><\/a><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/ad.linksynergy.com\/fs-bin\/show?id=d1QCig2q3qI&amp;bids=323058.950390&amp;type=2&amp;subid=0\" width=\"1\" height=\"1\" border=\"0\" \/><\/td>\n<td style=\"width: 203.15625px;\"><span class=\"rating-and-enrolled__element\">Students Enrolled::\u00a0<strong>19,359<\/strong><\/span><\/p>\n<p>Course Overall Rating::\u00a0<strong>4.6\u00a0<\/strong><\/td>\n<td style=\"width: 250px;\">\n<ul class=\"regular-list\">\n<li>Master Machine Learning on Python &amp; R<\/li>\n<li>Make robust Machine Learning models.<\/li>\n<li>Handle specific topics like Reinforcement Learning, NLP and Deep Learning.<\/li>\n<li>Build an army of powerful Machine Learning models and know how to combine them to solve any problem.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"width: 269.84375px;\"><a href=\"http:\/\/click.linksynergy.com\/link?id=d1QCig2q3qI&amp;offerid=479491.1560515878&amp;type=2&amp;murl=https%3A%2F%2Fwww.coursera.org%2Flearn%2Fml-classification\" target=\"_blank\">Machine Learning: Classification<\/a><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/ad.linksynergy.com\/fs-bin\/show?id=d1QCig2q3qI&amp;bids=479491.1560515878&amp;type=2&amp;subid=0\" width=\"1\" height=\"1\" border=\"0\" \/><a href=\"http:\/\/click.linksynergy.com\/link?id=d1QCig2q3qI&amp;offerid=479491.1560515878&amp;type=2&amp;murl=https%3A%2F%2Fwww.coursera.org%2Flearn%2Fml-classification\" target=\"_blank\"><img decoding=\"async\" data-src=\"https:\/\/d3njjcbhbojbot.cloudfront.net\/api\/utilities\/v1\/imageproxy\/https:\/\/coursera-course-photos.s3.amazonaws.com\/b1\/37a1a04d0011e59bd777c2a1ef6a6e\/gears-818461_1280.jpg\" border=\"0\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" class=\"lazyload\" \/><noscript><img decoding=\"async\" src=\"https:\/\/d3njjcbhbojbot.cloudfront.net\/api\/utilities\/v1\/imageproxy\/https:\/\/coursera-course-photos.s3.amazonaws.com\/b1\/37a1a04d0011e59bd777c2a1ef6a6e\/gears-818461_1280.jpg\" border=\"0\" \/><\/noscript><\/a><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/ad.linksynergy.com\/fs-bin\/show?id=d1QCig2q3qI&amp;bids=479491.1560515878&amp;type=2&amp;subid=0\" width=\"1\" height=\"1\" border=\"0\" \/><\/td>\n<td style=\"width: 203.15625px;\">Course Overall Rating::\u00a0<strong>4.7<\/strong><\/td>\n<td style=\"width: 250px;\">\n<ul>\n<li>Describe the input and output of a classification model.<\/li>\n<li>Build a classification model to predict sentiment in a product review dataset.<\/li>\n<li>Analyze financial data to predict loan defaults.<\/li>\n<\/ul>\n<ul>\n<li>Evaluate your models using precision-recall metrics.<\/li>\n<li>Implement these techniques in Python (or in the language of your choice, though Python is highly recommended).<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"width: 269.84375px;\"><a href=\"http:\/\/click.linksynergy.com\/link?id=d1QCig2q3qI&amp;offerid=479491.1560519184&amp;type=2&amp;murl=https%3A%2F%2Fwww.coursera.org%2Flearn%2Fpractical-machine-learning\" target=\"_blank\">Practical Machine Learning<\/a><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/ad.linksynergy.com\/fs-bin\/show?id=d1QCig2q3qI&amp;bids=479491.1560519184&amp;type=2&amp;subid=0\" width=\"1\" height=\"1\" border=\"0\" \/><br \/>\n<a href=\"http:\/\/click.linksynergy.com\/link?id=d1QCig2q3qI&amp;offerid=479491.1560519184&amp;type=2&amp;murl=https%3A%2F%2Fwww.coursera.org%2Flearn%2Fpractical-machine-learning\" target=\"_blank\"><img decoding=\"async\" data-src=\"https:\/\/d3njjcbhbojbot.cloudfront.net\/api\/utilities\/v1\/imageproxy\/https:\/\/coursera-course-photos.s3.amazonaws.com\/a1\/e7472069b611e3ae92c39913bb30e0\/PredictionMachineLearning.jpg\" border=\"0\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" class=\"lazyload\" \/><noscript><img decoding=\"async\" src=\"https:\/\/d3njjcbhbojbot.cloudfront.net\/api\/utilities\/v1\/imageproxy\/https:\/\/coursera-course-photos.s3.amazonaws.com\/a1\/e7472069b611e3ae92c39913bb30e0\/PredictionMachineLearning.jpg\" border=\"0\" \/><\/noscript><\/a><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/ad.linksynergy.com\/fs-bin\/show?id=d1QCig2q3qI&amp;bids=479491.1560519184&amp;type=2&amp;subid=0\" width=\"1\" height=\"1\" border=\"0\" \/><\/td>\n<td style=\"width: 203.15625px;\">&nbsp;<\/p>\n<p>Course Overall\u00a0Rating::\u00a0<strong>4.4<\/strong><\/td>\n<td style=\"width: 250px;\">\n<ul>\n<li>This course will cover the basic components of building and applying prediction functions with an emphasis on practical applications.<\/li>\n<li>Will provide the\u00a0basic grounding in concepts such as training and tests sets, overfitting, and error rates and introduce a range of model-based and algorithmic machine learning methods<\/li>\n<li>These methods including regression, classification trees, Naive Bayes, and random forests.<\/li>\n<li>The course will cover the complete process of building prediction functions including data collection, feature creation, algorithms, and evaluation.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>&nbsp;<\/p>\n","protected":false},"excerpt":{"rendered":"<p>SVM Classifier Introduction Hi, welcome to the another post on classification concepts. So far we have talked bout different classification concepts like logistic regression, knn classifier, decision trees .., etc. In this article, we were going to discuss \u00a0support vector machine which is a supervised learning algorithm. Just to give why we were so interested<br \/><a class=\"moretag\" href=\"https:\/\/dataaspirant.com\/support-vector-machine-algorithm\/\">+ Read More<\/a><\/p>\n","protected":false},"author":5,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":"","_jetpack_memberships_contains_paid_content":false},"categories":[3,4,6],"tags":[39,73],"aioseo_notices":[],"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/posts\/1239"}],"collection":[{"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/comments?post=1239"}],"version-history":[{"count":0,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/posts\/1239\/revisions"}],"wp:attachment":[{"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/media?parent=1239"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/categories?post=1239"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/tags?post=1239"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}