{"id":16295,"date":"2022-06-21T08:40:36","date_gmt":"2022-06-21T08:40:36","guid":{"rendered":"https:\/\/blog.datumo.com\/en\/?p=16295"},"modified":"2024-10-22T08:46:21","modified_gmt":"2024-10-22T08:46:21","slug":"deep-learning-the-basics-and-more","status":"publish","type":"post","link":"https:\/\/blog.datumo.com\/en\/tech\/16295","title":{"rendered":"Deep Learning, the basics and more!"},"content":{"rendered":"[vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1646799961152-e3ee06c0-4e82\" class=\"w-100 d-block \"><\/div><div class=\"pix-content-box card      vc_custom_1654577545529 custom-responsive-182427935   rounded-lg bg- w-100  \"   ><div class=\"\" style=\"z-index:30;position:relative;\">[vc_column_text]\r\n<p style=\"text-align: left;\"><span style=\"font-size: 14pt;\"><strong>\ud83d\udd11<\/strong> <strong>In 7 minutes you will learn:<\/strong><\/span><\/p>\r\n&nbsp;\r\n<ul>\r\n \t<li>The basics of deep learning<\/li>\r\n \t<li>The necessary steps to develop a model with deep neural network architecture<\/li>\r\n \t<li>Coding examples in Python 3<\/li>\r\n<\/ul>\r\n[\/vc_column_text]<\/div><\/div>[\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1650294698986-a1b962b5-ef42\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1655800989024{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: 40px !important;padding-left: 20px !important;}&#8221;]We will start off with its basics and then build on that to dive deeper into more complex structures. During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones. We will also see coding examples in Python 3 using popular frameworks like Tensorflow to get a better understanding of the concepts we cover. So, let\u2019s start![\/vc_column_text][\/vc_column][\/vc_row][vc_section full_width=&#8221;stretch_row&#8221; pix_over_visibility=&#8221;&#8221; css=&#8221;.vc_custom_1650444445523{padding-top: 80px !important;padding-bottom: 80px !important;background-color: #f8f9fa !important;}&#8221; el_id=&#8221;pix_section_program&#8221;][vc_row full_width=&#8221;stretch_row&#8221; pix_particles_check=&#8221;&#8221;][vc_column content_align=&#8221;text-center&#8221; offset=&#8221;vc_col-lg-offset-0 vc_col-lg-12 vc_col-md-offset-1 vc_col-md-10&#8243;][vc_raw_html]JTNDbWV0YSUyMGh0dHAtZXF1aXYlM0QlMjJyZWZyZXNoJTIyJTIwY29udGVudCUzRCUyMjAlM0IlMjB1cmwlM0RodHRwcyUzQSUyRiUyRmRhdHVtby5jb20lMkZlbiUyRmNyZWF0aW5nLXRoZS1iZXN0LXF1YWxpdHktaW1hZ2UtZGF0YXNldCUyRiUyMiUzRQ==[\/vc_raw_html]<div id=\"el1650442503491-f5da6b2f-fa35\" class=\"mb-3 text-left \"><h2 class=\"mb-32 pix-sliding-headline font-weight-bold secondary-font\" data-class=\"secondary-font text-heading-default\" data-style=\"\">Prerequisites<\/h2><\/div><div id=\"el1655802268158-d0122b5e-0f91\" class=\"w-100 d-block \"><\/div>[vc_column_text]\r\n<p style=\"text-align: left;\">To follow this article, you should have some prior basic knowledge of Machine Learning including Neural Networks, as well as basic programming knowledge in any language (preferably in Python). Other than that, the rest of the article is beginner-friendly.<\/p>\r\n[\/vc_column_text][\/vc_column][\/vc_row][\/vc_section][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1650442607008-a85a832d-43f0\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h2 class=\"text-heading-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">What\u2019s Deep Learning?<\/h2><\/div><\/div><\/div><div id=\"el1655802112138-15c06c3f-1a44\" class=\"w-100 d-block \"><\/div>[vc_column_text]Deep Learning, as stated previously, is a branch of Machine Learning which deals with handling complex datasets. The intriguing thing about it is that it uses the same methodology that our brain uses to learn new things i.e. through Neural Networks. Let us see what a Deep Neural Network looks like:\r\n\r\n&nbsp;\r\n\r\n<img fetchpriority=\"high\" decoding=\"async\" class=\"aligncenter size-full wp-image-16297\" src=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/DEEP.png\" alt=\"\" width=\"623\" height=\"418\" srcset=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/DEEP.png 623w, https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/DEEP-300x201.png 300w\" sizes=\"(max-width: 623px) 100vw, 623px\" \/>\r\n\r\n&nbsp;\r\n<p id=\"e2eb\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" data-selectable-paragraph=\"\">Scared? Don\u2019t be. We\u2019ll go through it part by part and through questions. So let us start with the first layer i.e.\u00a0<code class=\"hh ajr ajs ajt aig b\">Input Nodes<\/code>.<\/p>\r\n\r\n<h4><\/h4>\r\n&nbsp;\r\n<h4 class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\"><strong class=\"bn ml\">\r\nHow do we decide the number of neurons in the first layer?:<\/strong><\/h4>\r\n&nbsp;\r\n<p id=\"0e7c\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\">It depends on the features of your dataset. If you\u2019re feeding a text or numerical data in your network, the number of neurons would be equal to (the number of columns \u2014 1) of your dataset; the column that is skipped is the\u00a0<code class=\"hh ajr ajs ajt aig b\">Class<\/code>\u00a0or\u00a0<code class=\"hh ajr ajs ajt aig b\">Labels<\/code>\u00a0column that holds the values for each record that you wish for it to predict. Similarly, if you\u2019re feeding the image dataset to a neural net, then the number of neurons in the hidden layer would be equal to the dimensions of the images in the dataset. For instance, all your images are of dimension: 10 x 10. Then, the number of neurons would be 100.<\/p>\r\n&nbsp;\r\n\r\n&nbsp;\r\n<h4><strong class=\"bn ml\">What are hidden layers and how many layers do we add?:<\/strong><\/h4>\r\n&nbsp;\r\n\r\nHidden layers are there for the neural network\/ model to learn features in your dataset. You cannot know what feature each layer or neuron is learning, that is implicit. For a Deep Neural Network, there must be two or more hidden layers, otherwise, it would be considered a\u00a0<code class=\"hh ajr ajs ajt aig b\">shallow neural network<\/code>. Ok, so we have established that we need two or more layers, how do we decide the exact number? And the number of neurons in each of these hidden layers? Well, that is decided through the hit and trial method, or you can use\u00a0<code class=\"hh ajr ajs ajt aig b\">Grid Search<\/code>\u00a0algorithm to find the best values for you. If your dataset is small and the features it contains are simple, you should go with a smaller number of layers, if it&#8217;s large and complex, then the number of hidden layers needs to be increased as well as the number of neurons in each.[\/vc_column_text]<div id=\"el1650442651668-7359ff25-270a\" class=\"w-100 d-block \"><\/div><div id=\"el1650294913061-211813f5-5f2d\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row][vc_section full_width=&#8221;stretch_row&#8221; pix_over_visibility=&#8221;&#8221; css=&#8221;.vc_custom_1650444445523{padding-top: 80px !important;padding-bottom: 80px !important;background-color: #f8f9fa !important;}&#8221;][vc_row full_width=&#8221;stretch_row&#8221; pix_particles_check=&#8221;&#8221;][vc_column content_align=&#8221;text-center&#8221; offset=&#8221;vc_col-lg-offset-0 vc_col-lg-12 vc_col-md-offset-1 vc_col-md-10&#8243;]<div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h2 class=\"text-heading-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Steps to make a model with deep neural network architecture<\/h2><\/div><\/div><\/div><div id=\"el1655802279694-9cecf820-599f\" class=\"w-100 d-block \"><\/div>[vc_column_text]\r\n<p id=\"12c7\" class=\"pw-post-body-paragraph xt xu wo bn b xv xw hk xx xy xz ho ya yb yc yd ye yf yg yh yi yj yk yl ym yn jn iz\" style=\"text-align: left;\" data-selectable-paragraph=\"\">The next thing we need to learn is how to develop a model that makes uses of a deep neural network. The steps are mentioned below in sequence:<\/p>\r\n&nbsp;\r\n<ul class=\"\" style=\"text-align: left;\">\r\n \t<li id=\"4de3\" class=\"ahl ahm wo bn b xv yo xy yp yb ahn yf aho yj ahp yn akm ahr ahs aht iz\" data-selectable-paragraph=\"\">Load the dataset into your program<\/li>\r\n \t<li id=\"b5bb\" class=\"ahl ahm wo bn b xv ahu xy ahv yb ahw yf ahx yj ahy yn akm ahr ahs aht iz\" data-selectable-paragraph=\"\">Split your dataset into a training set and test set. The ratio is usually 70:30 i.e. you use 70% of your dataset for training and the remaining 30% for testing<\/li>\r\n \t<li id=\"a0bc\" class=\"ahl ahm wo bn b xv ahu xy ahv yb ahw yf ahx yj ahy yn akm ahr ahs aht iz\" data-selectable-paragraph=\"\">Preprocess your dataset to convert it into a format that can be fed into your deep neural network<\/li>\r\n \t<li id=\"2f75\" class=\"ahl ahm wo bn b xv ahu xy ahv yb ahw yf ahx yj ahy yn akm ahr ahs aht iz\" data-selectable-paragraph=\"\">Define your Deep Neural Network\u2019s architecture, i.e. how many layers it would have, the number of neurons in each layer, etc.<\/li>\r\n \t<li id=\"1bd1\" class=\"ahl ahm wo bn b xv ahu xy ahv yb ahw yf ahx yj ahy yn akm ahr ahs aht iz\" data-selectable-paragraph=\"\">Compile your model i.e. combine all the layers you have defined in the step above into a single model instance\/ object<\/li>\r\n \t<li id=\"6667\" class=\"ahl ahm wo bn b xv ahu xy ahv yb ahw yf ahx yj ahy yn akm ahr ahs aht iz\" data-selectable-paragraph=\"\">Train\/ fit your model on the training dataset<\/li>\r\n \t<li id=\"d22e\" class=\"ahl ahm wo bn b xv ahu xy ahv yb ahw yf ahx yj ahy yn akm ahr ahs aht iz\" data-selectable-paragraph=\"\">Evaluate your trained model on the testing set to see how it would perform on data that it has never seen before<\/li>\r\n<\/ul>\r\n&nbsp;\r\n<p id=\"6cef\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" style=\"text-align: left;\" data-selectable-paragraph=\"\">So, what are you waiting for? Let us jump right into it and get our hands dirty with some code.<\/p>\r\n[\/vc_column_text][\/vc_column][\/vc_row][\/vc_section][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1650362147064-486b7dc2-a9b3\" class=\"w-100 d-block \"><\/div><div id=\"el1650450433074-0be5e40e-928e\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h2 class=\"text-heading-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Coding Example<\/h2><\/div><\/div><\/div>[vc_column_text css=&#8221;.vc_custom_1655957601281{padding-top: 40px !important;padding-bottom: 0px !important;}&#8221;]\r\n<p id=\"1ceb\" class=\"pw-post-body-paragraph xt xu wo bn b xv xw hk xx xy xz ho ya yb yc yd ye yf yg yh yi yj yk yl ym yn jn iz\" data-selectable-paragraph=\"\">In this example, we are going to work on the\u00a0<a class=\"au mn\" href=\"https:\/\/archive.ics.uci.edu\/ml\/datasets\/Sentiment+Labelled+Sentences\" target=\"_blank\" rel=\"noopener ugc nofollow\">Sentiment Labelled Sentences Data Set<\/a>. You can download it by clicking.<\/p>\r\n&nbsp;\r\n<p id=\"7a87\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" data-selectable-paragraph=\"\">This is a dataset that contains user reviews from different websites including IMDB and yelp. In our example, we are only going to use the data from IMDB; it\u2019s very small so the accuracy will be poor, but we would be able to go through all the steps we defined above on this dataset. It has two columns, one is the content of the review, and the second column represents the overall sentiment of that comment, as either positive (1) or negative (0).\r\nSo, let\u2019s start!<\/p>\r\n&nbsp;\r\n<h6 id=\"4280\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\"><strong class=\"bn ml\">Note:<\/strong><\/h6>\r\n&nbsp;\r\n<p class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" data-selectable-paragraph=\"\">Our focus point here is the deep learning part, not the parts before it. We will cover the other steps, but we won\u2019t go into their detail. Furthermore, this tutorial is aimed towards getting you started on Deep Learning and give you a good idea of it, therefore, the minor parts would not be considered in too much detail.<\/p>\r\n&nbsp;\r\n<pre class=\"EnlighterJSRAW\" data-enlighter-language=\"generic\"># Step 1: Load the dataset\r\nimport pandas as pd \r\ndf = pd.read_csv ('imdb_labelled.txt', names= ['sentence', 'label'], sep = '\\t')<\/pre>\r\nNext, we will split our dataset into training and testing set with a 70 to 30 ration.\r\n\r\n&nbsp;\r\n<pre class=\"EnlighterJSRAW\" data-enlighter-language=\"generic\"># Step 2: Split dataset into training and test set\r\nprint(df.shape) # to see the number of records and columns\/ features in the dataset\r\nfrom sklearn.model_selection import train_test_split\r\nX = df['sentence'].values\r\nY = df['label'].values\r\nX_train, X_test, y_train, y_test = train_test_split(\r\n    sentences, y, test_size = 0.30, random_state = 10)<\/pre>\r\n&nbsp;\r\n<figure class=\"abo abp abq abr xj abs jc jd paragraph-image\"><\/figure>\r\n<p id=\"d07b\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" data-selectable-paragraph=\"\">Next, we will preprocess our dataset before feeding it into the model.<\/p>\r\n\r\n<pre class=\"EnlighterJSRAW\" data-enlighter-language=\"generic\"># Step 3: Preprocessing the dataset\r\nfrom sklearn.feature_extraction.text import CountVectorizer\r\nvectorizer = CountVectorizer()\r\nvectorizer.fit (X_train)\r\nX_train = vectorizer.transform (X_train)\r\nX_test  = vectorizer.transform (X_test)<\/pre>\r\n&nbsp;\r\n\r\n&nbsp;\r\n<p id=\"bee1\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" data-selectable-paragraph=\"\">And here comes the important part. We are now going to define our deep neural network\u2019s architecture, layer by layer, and then compile it.<\/p>\r\n\r\n<pre class=\"EnlighterJSRAW\" data-enlighter-language=\"generic\"># Step 4: Define the model\r\nfrom keras.models import Sequential\r\nfrom keras import layers\r\ninput_dim = X_train.shape[1]  # Number of features\r\nmodel = Sequential()\r\nmodel.add(layers.Dense(5, input_dim = input_dim, activation = 'relu'))\r\nmodel.add(layers.Dense(5, input_dim = input_dim, activation = 'relu'))\r\nmodel.add(layers.Dense(1, activation = 'sigmoid'))\r\n# since our output is just 0 or 1 to represent positive or negative \r\n# sentiment, we only have 1 neuron in the last\/ output layer\r\n# Step 5: Compiling the model\r\nmodel.compile(loss = 'binary_crossentropy', optimizer = 'adam', \r\n                metrics = ['accuracy'])\r\nmodel.summary()<\/pre>\r\n&nbsp;\r\n\r\n&nbsp;\r\n<h6><strong class=\"bn ml\">Output:<\/strong><\/h6>\r\n<table style=\"border-collapse: collapse; width: 100%;\">\r\n<tbody>\r\n<tr>\r\n<td style=\"width: 28.1215%; height: auto; vertical-align: middle;\"><strong>Layer (type)<\/strong><\/td>\r\n<td style=\"width: 41.17%; height: auto; vertical-align: middle;\"><strong>Output Shape<\/strong><\/td>\r\n<td style=\"width: 230.708%; height: auto; vertical-align: middle;\"><strong>Param #<\/strong><\/td>\r\n<\/tr>\r\n<tr style=\"height: 24px;\">\r\n<td style=\"width: 28.1215%; vertical-align: middle; height: 24px;\">dense_4 (Dense)<\/td>\r\n<td style=\"width: 41.17%; vertical-align: middle; height: 24px;\">(None, 5)<\/td>\r\n<td style=\"width: 230.708%; vertical-align: middle; height: 24px;\">12975<\/td>\r\n<\/tr>\r\n<tr style=\"height: 24px;\">\r\n<td style=\"width: 28.1215%; vertical-align: middle; height: 24px;\">dense_5 (Dense)<\/td>\r\n<td style=\"width: 41.17%; vertical-align: middle; height: 24px;\">(None, 5)<\/td>\r\n<td style=\"width: 230.708%; vertical-align: middle; height: 24px;\">30<\/td>\r\n<\/tr>\r\n<tr style=\"height: 24px;\">\r\n<td style=\"width: 28.1215%; vertical-align: middle; height: 24px;\">dense_6 (Dense)<\/td>\r\n<td style=\"width: 41.17%; vertical-align: middle; height: 24px;\">(None, 1)<\/td>\r\n<td style=\"width: 230.708%; vertical-align: middle; height: 24px;\">6<\/td>\r\n<\/tr>\r\n<tr style=\"height: 136px;\">\r\n<td style=\"width: 300%; height: 136px;\" colspan=\"3\">Total params: 13. 011\r\n\r\nTrainable params: 13, 011\r\n\r\nNon-trainable params: 0<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table>\r\n&nbsp;\r\n<p id=\"8dc3\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" data-selectable-paragraph=\"\">Now, we\u2019re headed towards the last two steps of our application. We will first fit our model onto the training set, and then evaluate it on our test set. So lets do that and see what\u2019s the accuracy we get on our testing set.<\/p>\r\n&nbsp;\r\n<pre class=\"EnlighterJSRAW\" data-enlighter-language=\"generic\"># Step 6: Fit the model on training set\r\nmodel.fit(X_train, y_train, epochs = 100, verbose = False,\r\n        validation_data = (X_test, y_test), batch_size = 10)\r\n# epoch means the number of times the model should iterate over the training set\r\n# Step 7: Evalute the model on testing set\r\nloss, accuracy = model.evaluate(X_test, y_test, verbose=False)\r\nprint(\"Testing Accuracy:  {:.2f}\".format(accuracy))<\/pre>\r\n&nbsp;\r\n<h6><strong class=\"bn ml\">Output:<\/strong><\/h6>\r\n&nbsp;\r\n<pre class=\"EnlighterJSRAW\" data-enlighter-language=\"generic\">Testing Accuracy:  0.78<\/pre>\r\n&nbsp;\r\n\r\nWell, accuracy is actually quite good. But you shouldn\u2019t expect that to be the case in your first try. The hyperparameters i.e. the number of layers and the number of neurons in each have currently been selected out of pure instinct and it is highly unlikely that these are the ones that give the best accuracy on this dataset. Usually, Grid Search is performed to find out the optimum hyperparameters. So, it\u2019s highly likely that if we applied that on our model, we would easily get accuracy above 80%. Pretty cool, right?[\/vc_column_text][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1653971463480-ce74a014-4ae9\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1655957611482{padding-top: 40px !important;padding-bottom: 0px !important;}&#8221;]\r\n<p id=\"9c25\" class=\"pw-post-body-paragraph xt xu wo bn b xv xw hk xx xy xz ho ya yb yc yd ye yf yg yh yi yj yk yl ym yn jn iz\" data-selectable-paragraph=\"\">Most times, it is always hard to gather enough data to have your own algorithm trained and tested. Therefore, it is often more efficient to find another service that does laborious works for you. We could be your perfect solution!<\/p>\r\n<p id=\"abd4\" class=\"pw-post-body-paragraph xt xu wo bn b xv yo hk xx xy yp ho ya yb yq yd ye yf yr yh yi yj ys yl ym yn jn iz\" data-selectable-paragraph=\"\">Here at <strong><a class=\"au mn\" href=\"https:\/\/www.datumo.com\" target=\"_blank\" rel=\"noopener ugc nofollow\"><em class=\"pn\">D<\/em><\/a><\/strong><a href=\"https:\/\/www.datumo.com\"><strong>ATUMO<\/strong><\/a>\u00a0, we crowdsource our tasks to diverse users located globally to ensure the quality and quantity on time. Moreover, our in-house managers double-check the quality of the collected or processed data. Let it be your professional dataset or academic dataset. We are here to help!<\/p>\r\n[\/vc_column_text]<div id=\"el1653972293756-76a5ecd1-3d25\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1655803104166{border-top-width: 1px !important;padding-top: 80px !important;padding-bottom: 0px !important;border-top-color: rgba(0,0,0,0.2) !important;border-top-style: solid !important;}&#8221;]To sum it all up, we started off by getting an introduction to what deep learning is and went on to discuss what we need to know before we jump into this domain. Later, we covered some common questions, the answers to which help us understand better what deep learning is. After that, we discussed the common steps in basic to intermediate Machine Learning applications that make use of deep learning. Lastly, we applied our knowledge and steps on a real dataset to make a model using a deep learning architecture.[\/vc_column_text]<div id=\"el1653971463481-f4f34d7c-39ce\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column width=&#8221;1\/2&#8243;]<div id=\"el1646794934167-c0c94dd3-ea74\" class=\"w-100 d-block \"><\/div><div class=\" mb-3 mb-md-0 \"  ><div class=\"card w-100 h-100 bg-white  vc_custom_1652982865548  pix-hover-item rounded-10 position-relative overflow-hidden2 text-white tilt fancy_card\" ><div class=\"card-img-overlay overflow-visible d-inline-block w-100 pix-img-overlay pix-p-30 d-flex align-items-end text-left\"><div class=\"w-100 \"><h3 class=\"card-title  text-black font-weight-bold mb-0 animate-in\" style=\"\">See what we can do for you.<\/h3><p class=\"card-text pix-pt-10 text-black \" style=\"\">Build smarter AI with us.<\/p><div class=\"card-btn-div mt-4 d-inline-block w-100\"><a  href=\"https:\/\/datumo.com\" class=\"btn mb-2     text-white btn-black d-inline-block      btn-md\" target=\"_blank\" rel=\"noopener\"    ><span class=\"font-weight-bold \" >Learn More<\/span><\/a><\/div><\/div><\/div><\/div><\/div>[\/vc_column][vc_column width=&#8221;1\/2&#8243;]<div id=\"el1646794982519-9a19190b-7fde\" class=\"w-100 d-block \"><\/div><div class=\" mb-3 mb-md-0 \"  ><div class=\"card w-100 h-100 bg-black  vc_custom_1653971438710  pix-hover-item rounded-10 position-relative overflow-hidden2 text-white tilt fancy_card\" ><div class=\"card-img-overlay overflow-visible d-inline-block w-100 pix-img-overlay pix-p-30 d-flex align-items-end text-left\"><div class=\"w-100 \"><h3 class=\"card-title  text-white font-weight-bold mb-0 animate-in\" style=\"\">We would like to support the AI industry by sharing.<\/h3><p class=\"card-text pix-pt-10 text-white \" style=\"\"><\/p><div class=\"card-btn-div mt-4 d-inline-block w-100\"><a  href=\"https:\/\/open.datumo.com\/en\" class=\"btn mb-2    vc_custom_1653971438714  btn-primary d-inline-block      btn-md\" target=\"_blank\" rel=\"noopener\"    ><span class=\"font-weight-bold \" >Download Open Datasets<\/span><\/a><\/div><\/div><\/div><\/div><\/div>[\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1646799961152-e3ee06c0-4e82\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row]","protected":false},"excerpt":{"rendered":"[vc_row pix_particles_check=&#8221;&#8221;][vc_column][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column][vc_column_text css=&#8221;.vc_custom_1655800989024{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: 40px !important;padding-left: 20px !important;}&#8221;]We will start off with its basics and then build on that to dive deeper into more complex structures. During the transition from simple to complex, we will discuss&#8230;","protected":false},"author":1,"featured_media":16449,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[131],"tags":[130,127,182,150,183],"class_list":["post-16295","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tech","tag-datasets","tag-datumo","tag-deep-learning","tag-machine-learning","tag-python3"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v23.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Deep Learning, the basics and more! - DATUMO<\/title>\n<meta name=\"description\" content=\"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/blog.datumo.com\/en\/tech\/16295\" \/>\n<meta property=\"og:locale\" content=\"ko_KR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Deep Learning, the basics and more!\" \/>\n<meta property=\"og:description\" content=\"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/blog.datumo.com\/en\/tech\/16295\" \/>\n<meta property=\"og:site_name\" content=\"DATUMO\" \/>\n<meta property=\"article:published_time\" content=\"2022-06-21T08:40:36+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-10-22T08:46:21+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1574\" \/>\n\t<meta property=\"og:image:height\" content=\"715\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"DATUMO\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:title\" content=\"Deep Learning, the basics and more!\" \/>\n<meta name=\"twitter:description\" content=\"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png\" \/>\n<meta name=\"twitter:label1\" content=\"\uae00\uc4f4\uc774\" \/>\n\t<meta name=\"twitter:data1\" content=\"DATUMO\" \/>\n\t<meta name=\"twitter:label2\" content=\"\uc608\uc0c1 \ub418\ub294 \ud310\ub3c5 \uc2dc\uac04\" \/>\n\t<meta name=\"twitter:data2\" content=\"9\ubd84\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"TechArticle\",\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295#article\",\"isPartOf\":{\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295\"},\"author\":{\"name\":\"DATUMO\",\"@id\":\"https:\/\/blog.datumo.com\/#\/schema\/person\/02ec2d0ba953b146878dab089dc735b6\"},\"headline\":\"Deep Learning, the basics and more!\",\"datePublished\":\"2022-06-21T08:40:36+00:00\",\"dateModified\":\"2024-10-22T08:46:21+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295\"},\"wordCount\":2159,\"publisher\":{\"@id\":\"https:\/\/blog.datumo.com\/#organization\"},\"image\":{\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage\"},\"thumbnailUrl\":\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png\",\"keywords\":[\"datasets\",\"datumo\",\"deep learning\",\"machine learning\",\"python3\"],\"articleSection\":[\"tech\"],\"inLanguage\":\"ko-KR\"},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295\",\"url\":\"https:\/\/blog.datumo.com\/en\/tech\/16295\",\"name\":\"Deep Learning, the basics and more! - DATUMO\",\"isPartOf\":{\"@id\":\"https:\/\/blog.datumo.com\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage\"},\"image\":{\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage\"},\"thumbnailUrl\":\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png\",\"datePublished\":\"2022-06-21T08:40:36+00:00\",\"dateModified\":\"2024-10-22T08:46:21+00:00\",\"description\":\"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.\",\"breadcrumb\":{\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295#breadcrumb\"},\"inLanguage\":\"ko-KR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/blog.datumo.com\/en\/tech\/16295\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage\",\"url\":\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png\",\"contentUrl\":\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png\",\"width\":1574,\"height\":715},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/blog.datumo.com\/en\/tech\/16295#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/blog.datumo.com\/en\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Deep Learning, the basics and more!\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/blog.datumo.com\/#website\",\"url\":\"https:\/\/blog.datumo.com\/\",\"name\":\"DATUMO\",\"description\":\"The Data for Smarter AI\",\"publisher\":{\"@id\":\"https:\/\/blog.datumo.com\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/blog.datumo.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"ko-KR\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/blog.datumo.com\/#organization\",\"name\":\"DATUMO\",\"url\":\"https:\/\/blog.datumo.com\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/blog.datumo.com\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/05\/2.1.webp\",\"contentUrl\":\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/05\/2.1.webp\",\"width\":1080,\"height\":600,\"caption\":\"DATUMO\"},\"image\":{\"@id\":\"https:\/\/blog.datumo.com\/#\/schema\/logo\/image\/\"}},{\"@type\":\"Person\",\"@id\":\"https:\/\/blog.datumo.com\/#\/schema\/person\/02ec2d0ba953b146878dab089dc735b6\",\"name\":\"DATUMO\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/blog.datumo.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/1942a8a63e1c8fa0d9be56cda789edd6c0a866259cd5dca24952597ffa8bab3d?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/1942a8a63e1c8fa0d9be56cda789edd6c0a866259cd5dca24952597ffa8bab3d?s=96&d=mm&r=g\",\"caption\":\"DATUMO\"},\"description\":\"DATUMO, The Data for Smarter AI. We seek to drive impact in the world by providing diverse and high quality data to build smarter AI.\",\"sameAs\":[\"https:\/\/blog.datumo.com\/en\"],\"url\":\"https:\/\/blog.datumo.com\/en\/author\/selectstar\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Deep Learning, the basics and more! - DATUMO","description":"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.","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:\/\/blog.datumo.com\/en\/tech\/16295","og_locale":"ko_KR","og_type":"article","og_title":"Deep Learning, the basics and more!","og_description":"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.","og_url":"https:\/\/blog.datumo.com\/en\/tech\/16295","og_site_name":"DATUMO","article_published_time":"2022-06-21T08:40:36+00:00","article_modified_time":"2024-10-22T08:46:21+00:00","og_image":[{"width":1574,"height":715,"url":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png","type":"image\/png"}],"author":"DATUMO","twitter_card":"summary_large_image","twitter_title":"Deep Learning, the basics and more!","twitter_description":"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.","twitter_image":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png","twitter_misc":{"\uae00\uc4f4\uc774":"DATUMO","\uc608\uc0c1 \ub418\ub294 \ud310\ub3c5 \uc2dc\uac04":"9\ubd84"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"TechArticle","@id":"https:\/\/blog.datumo.com\/en\/tech\/16295#article","isPartOf":{"@id":"https:\/\/blog.datumo.com\/en\/tech\/16295"},"author":{"name":"DATUMO","@id":"https:\/\/blog.datumo.com\/#\/schema\/person\/02ec2d0ba953b146878dab089dc735b6"},"headline":"Deep Learning, the basics and more!","datePublished":"2022-06-21T08:40:36+00:00","dateModified":"2024-10-22T08:46:21+00:00","mainEntityOfPage":{"@id":"https:\/\/blog.datumo.com\/en\/tech\/16295"},"wordCount":2159,"publisher":{"@id":"https:\/\/blog.datumo.com\/#organization"},"image":{"@id":"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage"},"thumbnailUrl":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png","keywords":["datasets","datumo","deep learning","machine learning","python3"],"articleSection":["tech"],"inLanguage":"ko-KR"},{"@type":"WebPage","@id":"https:\/\/blog.datumo.com\/en\/tech\/16295","url":"https:\/\/blog.datumo.com\/en\/tech\/16295","name":"Deep Learning, the basics and more! - DATUMO","isPartOf":{"@id":"https:\/\/blog.datumo.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage"},"image":{"@id":"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage"},"thumbnailUrl":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png","datePublished":"2022-06-21T08:40:36+00:00","dateModified":"2024-10-22T08:46:21+00:00","description":"During the transition from simple to complex, we will discuss why the need arises for the use of deep networks, and the shortcomings of the simple\/ naive ones.","breadcrumb":{"@id":"https:\/\/blog.datumo.com\/en\/tech\/16295#breadcrumb"},"inLanguage":"ko-KR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/blog.datumo.com\/en\/tech\/16295"]}]},{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/blog.datumo.com\/en\/tech\/16295#primaryimage","url":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png","contentUrl":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/06\/dasda\u3134\u3134.png","width":1574,"height":715},{"@type":"BreadcrumbList","@id":"https:\/\/blog.datumo.com\/en\/tech\/16295#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/blog.datumo.com\/en\/"},{"@type":"ListItem","position":2,"name":"Deep Learning, the basics and more!"}]},{"@type":"WebSite","@id":"https:\/\/blog.datumo.com\/#website","url":"https:\/\/blog.datumo.com\/","name":"DATUMO","description":"The Data for Smarter AI","publisher":{"@id":"https:\/\/blog.datumo.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/blog.datumo.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"ko-KR"},{"@type":"Organization","@id":"https:\/\/blog.datumo.com\/#organization","name":"DATUMO","url":"https:\/\/blog.datumo.com\/","logo":{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/blog.datumo.com\/#\/schema\/logo\/image\/","url":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/05\/2.1.webp","contentUrl":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/05\/2.1.webp","width":1080,"height":600,"caption":"DATUMO"},"image":{"@id":"https:\/\/blog.datumo.com\/#\/schema\/logo\/image\/"}},{"@type":"Person","@id":"https:\/\/blog.datumo.com\/#\/schema\/person\/02ec2d0ba953b146878dab089dc735b6","name":"DATUMO","image":{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/blog.datumo.com\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/1942a8a63e1c8fa0d9be56cda789edd6c0a866259cd5dca24952597ffa8bab3d?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/1942a8a63e1c8fa0d9be56cda789edd6c0a866259cd5dca24952597ffa8bab3d?s=96&d=mm&r=g","caption":"DATUMO"},"description":"DATUMO, The Data for Smarter AI. We seek to drive impact in the world by providing diverse and high quality data to build smarter AI.","sameAs":["https:\/\/blog.datumo.com\/en"],"url":"https:\/\/blog.datumo.com\/en\/author\/selectstar"}]}},"_links":{"self":[{"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/posts\/16295","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/comments?post=16295"}],"version-history":[{"count":16,"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/posts\/16295\/revisions"}],"predecessor-version":[{"id":16917,"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/posts\/16295\/revisions\/16917"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/media\/16449"}],"wp:attachment":[{"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/media?parent=16295"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/categories?post=16295"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.datumo.com\/en\/wp-json\/wp\/v2\/tags?post=16295"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}