{"id":3032,"date":"2017-05-03T02:51:35","date_gmt":"2017-05-03T02:51:35","guid":{"rendered":"http:\/\/dataaspirant.com\/?p=3032"},"modified":"2017-05-03T15:35:37","modified_gmt":"2017-05-03T15:35:37","slug":"handwritten-digits-recognition-tensorflow-python","status":"publish","type":"post","link":"https:\/\/dataaspirant.com\/handwritten-digits-recognition-tensorflow-python\/","title":{"rendered":"Handwritten digits recognition using google tensorflow with python"},"content":{"rendered":"<div id=\"bsf_rt_marker\"><\/div><div id=\"attachment_3035\" style=\"width: 710px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-3035\" class=\"size-full wp-image-3035 lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Handwritten-digits-recognition-using-tensorflow.jpg\" alt=\"Handwritten digits recognition using tensorflow\" width=\"700\" height=\"400\" data-srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Handwritten-digits-recognition-using-tensorflow.jpg 700w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Handwritten-digits-recognition-using-tensorflow-300x171.jpg 300w\" data-sizes=\"(max-width: 700px) 100vw, 700px\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 700px; --smush-placeholder-aspect-ratio: 700\/400;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-3035\" class=\"size-full wp-image-3035\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Handwritten-digits-recognition-using-tensorflow.jpg\" alt=\"Handwritten digits recognition using tensorflow\" width=\"700\" height=\"400\" srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Handwritten-digits-recognition-using-tensorflow.jpg 700w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Handwritten-digits-recognition-using-tensorflow-300x171.jpg 300w\" sizes=\"(max-width: 700px) 100vw, 700px\" \/><\/noscript><p id=\"caption-attachment-3035\" class=\"wp-caption-text\">Handwritten digits recognition using tensorflow<\/p><\/div>\n<h1>Handwritten digits recognition using Tensorflow with Python<\/h1>\n<p style=\"text-align: justify;\">The progress in technology that has happened over the last 10 years is unbelievable. Every corner of the world is using the top most technologies to improve existing products while also conducting immense research into inventing products that make the world the best place to live.<\/p>\n<p style=\"text-align: justify;\">Some of these are the <a href=\"https:\/\/dataaspirant.com\/2016\/12\/14\/amazon-go-just-walkout-technology\/\" target=\"_blank\">Amazon just walk out technology<\/a>, the Tesla autopilot car, Spaceships and more. All of these breakthrough products could never exist without\u00a0<a href=\"https:\/\/dataaspirant.com\/category\/machine-learning-2\/\" target=\"_blank\">machine learning<\/a> and <strong>deep learning algorithms<\/strong>.<\/p>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">Despite the complexity of the computations involved, some very sophisticated calculations can easily be handled by frameworks created for machine learning and deep learning. This article will help you get started with one of the most popular frameworks &#8211; Tensorflow.<\/span><\/p>\n<p style=\"text-align: justify;\">So in this article, you will get a taste of deep learning with some interesting application, the handwritten<strong> digits recognization<\/strong> application.<\/p>\n<p style=\"text-align: justify;\"><span class='bctt-click-to-tweet'><span class='bctt-ctt-text'><a href='https:\/\/twitter.com\/intent\/tweet?url=https%3A%2F%2Fdataaspirant.com%2F%3Fp%3D3032&#038;text=Handwritten%20digits%20recognition%20using%20google%20tensorflow%20with%20python&#038;via=dataaspirant&#038;related=dataaspirant' target='_blank'rel=\"noopener noreferrer\">Handwritten digits recognition using google tensorflow with python <\/a><\/span><a href='https:\/\/twitter.com\/intent\/tweet?url=https%3A%2F%2Fdataaspirant.com%2F%3Fp%3D3032&#038;text=Handwritten%20digits%20recognition%20using%20google%20tensorflow%20with%20python&#038;via=dataaspirant&#038;related=dataaspirant' target='_blank' class='bctt-ctt-btn'rel=\"noopener noreferrer\">Click To Tweet<\/a><\/span><\/p>\n<p style=\"text-align: justify;\">Before we begin. We would like to <strong>thank<\/strong>\u00a0Google for access to their open source the tensorflow library.<\/p>\n<h3>Table of contents:<\/h3>\n<ul>\n<li>What is Tensorflow?<\/li>\n<li>About the MNIST dataset<\/li>\n<li>Implementing the Handwritten digits recognition model<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<h2>What is Tensorflow?<\/h2>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">Tensorflow is an open source library created by the Google Brain Trust for heavy computational work, geared towards machine learning and deep learning tasks. It is built on C, C++ making its computations very fast while it is available for use via a Python, C++, Haskell, Java and Go API. <\/span><\/p>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">It created data graph flows for each model, where a graph consists of two units &#8211; a <\/span><b>tensor<\/b><span style=\"font-weight: 400;\"> and a <\/span><b>node<\/b><span style=\"font-weight: 400;\">.\u00a0<\/span><\/p>\n<ul>\n<li style=\"text-align: justify;\"><strong>Tensor:<\/strong> A tensor is any multidimensional array.<\/li>\n<li style=\"text-align: justify;\"><strong>Node:<\/strong> A node is a mathematical computation that is being worked at the moment.<\/li>\n<\/ul>\n<p style=\"text-align: justify;\">A data graph flow essentially maps the flow of information via the interchange between these two components. Once this graph is complete, the model is executed and the output is computed.<\/p>\n<p style=\"text-align: justify;\">You can learn a lot more from the <a href=\"https:\/\/www.tensorflow.org\" target=\"_blank\">tensorflow official document<\/a><\/p>\n<p style=\"text-align: justify;\">Now let&#8217;s begin start building handwritten digits recognition application. To start we need the dataset of handwritten digits for training and for testing the model. \u00a0MNIST is the most popular dataset having handwritten digits as image files.<\/p>\n<h2>About the MNIST dataset<\/h2>\n<div id=\"attachment_3037\" style=\"width: 540px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-3037\" class=\"wp-image-3037 size-full lazyload\" data-src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Mnist-database-hand-written-digits.png\" alt=\"Mnist database handwritten digits\" width=\"530\" height=\"297\" data-srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Mnist-database-hand-written-digits.png 530w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Mnist-database-hand-written-digits-300x168.png 300w\" data-sizes=\"(max-width: 530px) 100vw, 530px\" src=\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\" style=\"--smush-placeholder-width: 530px; --smush-placeholder-aspect-ratio: 530\/297;\" \/><noscript><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-3037\" class=\"wp-image-3037 size-full\" src=\"http:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Mnist-database-hand-written-digits.png\" alt=\"Mnist database handwritten digits\" width=\"530\" height=\"297\" srcset=\"https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Mnist-database-hand-written-digits.png 530w, https:\/\/dataaspirant.com\/wp-content\/uploads\/2017\/05\/Mnist-database-hand-written-digits-300x168.png 300w\" sizes=\"(max-width: 530px) 100vw, 530px\" \/><\/noscript><p id=\"caption-attachment-3037\" class=\"wp-caption-text\">Mnist database handwritten digits<\/p><\/div>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">To begin our journey with Tensorflow, we will be using the MNIST database to create an image identifying model based on simple <strong>feedforward neural network<\/strong> with no hidden layers.<\/span><\/p>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">MNIST is a computer vision database consisting of handwritten digits, with labels identifying the digits. As mentioned earlier, every <strong>MNIST<\/strong> data point has two parts: an image of a handwritten digit and a corresponding label.\u00a0<\/span><\/p>\n<p style=\"text-align: justify;\">We&#8217;ll call the images <strong>&#8220;x&#8221;<\/strong> and the labels <strong>&#8220;y&#8221;<\/strong>. Both the training set and test set contain images and their corresponding labels; for example, the training images are <b>mnist.train.images <\/b>and the training labels are <b>mnist.train.labels<\/b>.<\/p>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">Each image is <strong>28<\/strong> pixels by <strong>28<\/strong> pixels. We can interpret this as a big array of numbers. We can flatten this array into a vector of 28&#215;28 = 784 numbers.\u00a0<\/span><\/p>\n<p style=\"text-align: justify;\">It doesn&#8217;t matter how we flatten the array, as long as we&#8217;re consistent between images. From this perspective, the MNIST images are just a bunch of points in a <strong>784<\/strong>-dimentional vector space.<\/p>\n<h2>Implementing the Handwritten digits recognition model<\/h2>\n<h3>Implementing the handwritten digits model using Tensorflow with Python<\/h3>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">We will be building simple feedforward neural network using <a href=\"https:\/\/dataaspirant.com\/2017\/03\/07\/difference-between-softmax-function-and-sigmoid-function\/\" target=\"_blank\">softmax<\/a> to predict the number in each image. We begin by calling in a Python environment.\u00a0<\/span><\/p>\n<p style=\"text-align: justify;\">As you need python as a prerequisite for understanding the below implementation. If you\u00a0are new to the python and facing any environment issues then get quick hands on experience on <a href=\"http:\/\/dataaspirant.com\/recommends\/data-science-courses\/byteacademy-software-development\/\" target=\"_blank\">python and the environment setup<\/a> before you start.<\/p>\n<h4><span style=\"font-size: 14pt;\"><strong>Download the MNIS<\/strong><\/span><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Download the dataset\">from tensorflow.examples.tutorials.mnist import input_data\r\nmnist = input_data.read_data_sets(\u201cmodel_data\/\", one_hot=True)<\/pre>\n<h4><strong>Import Tensorflow to your environment<\/strong><\/h4>\n<pre class=\"lang:python decode:true\" title=\"Import tensorflow\">import tensorflow as tf<\/pre>\n<h4><strong>Initializing parameters for the model<\/strong><\/h4>\n<pre class=\"lang:python decode:true\" title=\"Initializing parameters for the model\">batch =100\r\nlearning_rate=0.01\r\ntraining_epochs=10<\/pre>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">In machine learning, an epoch is a full iteration over samples. Here, we are restricting the model to 10 complete epochs or cycles of the algorithm running through the dataset.<\/span><\/p>\n<p style=\"text-align: justify;\">The\u00a0<span style=\"font-weight: 400;\">batch variable determines the amount of data being fed to the algorithm at any given time, in this case, 100 images.\u00a0<\/span><\/p>\n<p style=\"text-align: justify;\">The learning rate controls the size of the parameters and rates, thereby affecting the rate at which the model \u201clearns&#8221;.<\/p>\n<h4 style=\"text-align: justify;\"><strong>Creating Placeholders<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Creating Placeholders\">x = tf.placeholder(tf.float32, shape=[None, 784]) \r\ny_ = tf.placeholder(tf.float32, shape=[None, 10])<\/pre>\n<p>&nbsp;<\/p>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">The method tf.placeholder allows us to create variables that act as nodes holding the data. Here, x is a 2-dimensionall array holding the MNIST images, with none implying the batch size (which can be of any size) and 784 being a single 28&#215;28 image. y_ is the target output class that consists of a 2-dimensional array of 10 classes (denoting the numbers 0-9) that identify what digit is stored in each image.<\/span><\/p>\n<h4 style=\"text-align: justify;\"><strong>Creating Variables<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Creating Variables\">W = tf.Variable(tf.zeros([784, 10]))\r\nb = tf.Variable(tf.zeros([10]))\r\n<\/pre>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">Here, W is the <strong>weight<\/strong> and b is the <strong>bias<\/strong> of the model. They are initialized with <strong>tf.Variable<\/strong> as they are components of the computational graph that need to change values with the input of each different neuron.<\/span><\/p>\n<h4 style=\"text-align: justify;\"><strong>Initializing the model<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Initializing the model\">y = tf.nn.softmax(tf.matmul(x,W) + b)<\/pre>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">We will be using a simple softmax model to implement our network. Softmax is a generalization of logistic regression, usually used in the final layer of a network. It is useful because it helps in multi-classification models where a given output can be a list of many different things.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">It provides values between 0 to 1 that in addition give you the probability of the output belonging to a particular class. \u00a0<\/span><\/p>\n<h4><strong>Defining Cost Function<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Defining Cost Function\">cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y), reduction_indices=[1]))\r\n<\/pre>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">This is the cost function of the model &#8211; a cost function is a difference between the predicted value and the actual value that we are trying to minimize to improve the accuracy of the model.<\/span><\/p>\n<h4 style=\"text-align: justify;\"><strong>Determining the accuracy of parameters<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Determining the accuracy of parameters\">correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))\r\naccuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))<\/pre>\n<h4><strong>Implementing Gradient Descent \u00a0Algorithm<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Implementing Gradient Descent Algorithms\">train_op = tf.train.GradientDescentOptimizer(learning_rate).minimize(cross_entropy)<\/pre>\n<p><span style=\"font-weight: 400;\">Tensorflow comes pre-loaded with a lot of algorithms, one of them being Gradient Descent. The gradient descent algorithm starts with an initial value and keeps updating the value till the cost function reaches the global minimum i.e. the highest level of accuracy.\u00a0<\/span><\/p>\n<p>This is obviously dependant upon the number of iterations being permitted for the model.<\/p>\n<h4><strong>Initializing the session<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Initializing the session\">with tf.Session() as sess:\r\n  sess.run(tf.initialize_all_variables())<\/pre>\n<h4><strong>Creating batches of data for epochs<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Creating batches of data for epochs\">for epoch in range(training_epochs):\r\n    batch_count = int(mnist.train.num_examples\/batch)\r\n    for i in range(batch_count):\r\n        batch_x, batch_y = mnist.train.next_batch(batch)\r\n\r\n<\/pre>\n<h4><strong>Executing the model<\/strong><\/h4>\n<pre class=\"lang:python decode:true \" title=\"Executing the model\">sess.run([train_op], feed_dict={x: batch_x, y_: batch_y})<\/pre>\n<h4><strong>Print accuracy of the model<\/strong><\/h4>\n<pre class=\"lang:python decode:true\" title=\"Print accuracy of the model\">if epoch % 2 == 0: \r\n      print \"Epoch: \", epoch \r\n  prit \"Accuracy: \", accuracy.eval(feed_dict={x: mnist.test.images, y_: mnist.test.labels})\r\n  print \"Model Execution Complete\"<\/pre>\n<p>Share your accuracy details I would love to see those.<\/p>\n<h3>Complete Code:<\/h3>\n<pre class=\"lang:python decode:true  \" title=\"Complete code\">#!\/usr\/bin\/env python\r\n# handwritten_digits_recognition.py\r\n# Date: 03-May-2017\r\n# About: Handwritten digits recognition with Tensorflow\r\n\r\n\r\n# Required Python Packages\r\nfrom tensorflow.examples.tutorials.mnist import input_data\r\n\r\n# Download the MNIS dataset\r\nmnist = input_data.read_data_sets(\u201cmodel_data\/\", one_hot=True)\r\n\r\n# import tensorflow to the environment\r\nimport tensorflow as tf\r\n\r\n# initializing parameters for the model\r\nbatch = 100\r\nlearning_rate = 0.01\r\ntraining_epochs = 10\r\n\r\n# creating placeholders\r\nx = tf.placeholder(tf.float32, shape=[None, 784])\r\ny_ = tf.placeholder(tf.float32, shape=[None, 10])\r\n\r\n# creating variables\r\nW = tf.Variable(tf.zeros([784, 10]))\r\nb = tf.Variable(tf.zeros([10]))\r\n\r\n# initializing the model\r\ny = tf.nn.softmax(tf.matmul(x,W) + b)\r\n\r\n# Defining Cost Function\r\ncross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y), reduction_indices=[1]))\r\n\r\n# Determining the accuracy of parameters\r\ncorrect_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))\r\naccuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\r\n\r\n# Implementing Gradient Descent Algorithm\r\ntrain_op = tf.train.GradientDescentOptimizer(learning_rate).minimize(cross_entropy)\r\n\r\n# Initializing the session\r\nwith tf.Session() as sess:\r\n  sess.run(tf.initialize_all_variables())\r\n\r\n# Creating batches of data for epochs\r\nfor epoch in range(training_epochs):\r\n    batch_count = int(mnist.train.num_examples \/ batch)\r\n    for i in range(batch_count):\r\n        batch_x, batch_y = mnist.train.next_batch(batch)\r\n\r\n# Executing the model\r\nsess.run([train_op], feed_dict={x: batch_x, y_: batch_y})\r\n\r\n# Print Accuracy of the model\r\nif epoch % 2 == 0:\r\n    print \"Epoch: \", epoch\r\n    prnt \"Accuracy: \", accuracy.eval(feed_dict={x: mnist.test.images, y_: mnist.test.labels})\r\n    print \"Model Execution Complete\"\r\n\r\n\r\n\r\n\r\n<\/pre>\n<p>You can clone the complete code from our <a href=\"https:\/\/github.com\/saimadhu-polamuri\/DataAspirant_codes\/tree\/master\/handwritten_digits_recognition\" target=\"_blank\" rel=\"noopener noreferrer\">Github account\u00a0<\/a><\/p>\n<h3>Final Note<\/h3>\n<p style=\"text-align: justify;\"><span style=\"font-weight: 400;\">Creating a deep learning model can be easy and intuitive on Tensorflow. But to really implement some cool things, you need to have a good grasp on <a href=\"http:\/\/dataaspirant.com\/recommends\/data-science-courses\/byteacademy-data-science\/\" target=\"_blank\">machine learning principles used in data science<\/a>.<\/span><\/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>|\u00a0<span style=\"color: #808000;\"><a style=\"color: #808000;\" href=\"https:\/\/plus.google.com\/104087213513583475222\" target=\"_blank\">GOOGLE+<\/a><\/span>\u00a0|\u00a0<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>|\u00a0<span style=\"color: #ff0000;\"><a style=\"color: #ff0000;\" title=\"flipboard dataaspirant\" href=\"https:\/\/flipboard.com\/@dataaspirant\/\" target=\"_blank\">FLIPBOARD<\/a><\/span>\u00a0|\u00a0<span style=\"color: #00ff00;\"><a style=\"color: #00ff00;\" title=\"medium dataaspirant\" href=\"https:\/\/medium.com\/@dataaspirant\" target=\"_blank\">MEDIUM<\/a>\u00a0|\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","protected":false},"excerpt":{"rendered":"<p>Handwritten digits recognition using Tensorflow with Python The progress in technology that has happened over the last 10 years is unbelievable. Every corner of the world is using the top most technologies to improve existing products while also conducting immense research into inventing products that make the world the best place to live. Some of<br \/><a class=\"moretag\" href=\"https:\/\/dataaspirant.com\/handwritten-digits-recognition-tensorflow-python\/\">+ Read More<\/a><\/p>\n","protected":false},"author":6,"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":[1434],"tags":[1454],"aioseo_notices":[],"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/posts\/3032"}],"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\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/comments?post=3032"}],"version-history":[{"count":0,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/posts\/3032\/revisions"}],"wp:attachment":[{"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/media?parent=3032"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/categories?post=3032"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dataaspirant.com\/wp-json\/wp\/v2\/tags?post=3032"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}