{"id":1642,"date":"2020-09-17T03:44:21","date_gmt":"2020-09-17T03:44:21","guid":{"rendered":"https:\/\/data-science.gotoauthority.com\/2020\/09\/17\/introduction-to-neural-network-basics\/"},"modified":"2020-09-17T03:44:21","modified_gmt":"2020-09-17T03:44:21","slug":"introduction-to-neural-network-basics","status":"publish","type":"post","link":"https:\/\/wealthrevelation.com\/data-science\/2020\/09\/17\/introduction-to-neural-network-basics\/","title":{"rendered":"Introduction to Neural Network Basics"},"content":{"rendered":"
\n

This is the first part of a series of blog posts on simple Neural Networks. The basics of neural networks can be found all over the internet. Many of them are the same, each article is written slightly differently.\u00a0<\/p>\n

But here we tried a different approach to get a deep understanding of the neural networks by explaining\u00a0each building block<\/strong> concept to build the neural network.<\/p>\n

Literally, we will narrow down to the very basic concepts you should need to build the neural networks. The knowledge you gained in this article will help you understand the various deep learning models architecture<\/strong> in the long run.\u00a0<\/p>\n<\/div>\n

\n

Many of us have seen the pocket calculator<\/strong> in an arithmetic contest. It will never improve its speed or accuracy, no matter how much it practices.\u00a0<\/p>\n

In short:<\/h4>\n

It doesn’t learn.<\/strong><\/p><\/blockquote>\n

For example, every time I press its square-root button, it computes exactly the same function in exactly the same way. Here the pocket calculator is not learning.<\/p>\n

But how can it learn?\u00a0<\/p>\n

By computing a function. Our brains can also learn much more efficiently based on the same idea. Before delving deeper into how such networks can learn, let’s first understand how they can compute.<\/p>\n

This computing function is called neural networks models in deep learning, in machine learning literature it\u2019s called a machine learning model<\/a>.<\/p>\n

Unlike various machine learning models such as logistic regression<\/a>, decision trees<\/a>, randomforest<\/a> the deep learning models are complete different<\/strong> in the way they learn from data.<\/p>\n

Now let\u2019s learn how the neural networks learn from the data we are feeding.<\/p>\n

Introduction to Neural networks<\/h2>\n

A neural network is simply a group of interconnected neurons<\/strong> that are able to influence each other\u2019s behavior.\u00a0<\/p>\n

Your brain contains about as many neurons as there are stars in our galaxy. On average, each of these neurons is connected to a thousand other neurons via junctions called synapses<\/strong>.<\/p>\n

We can schematically draw a neural network as a collection of dots representing neurons connected by lines representing synapses as shown in the below figure.\u00a0<\/p>\n<\/div>\n

\n

Real-world neurons are very complicated. However, AI researchers have shown that neural networks can still attain human-level<\/strong> performance on many remarkably complex tasks.\u00a0<\/p>\n

Such as hand written text recognition<\/a>, identifiying cancer tumers ..etc<\/p>\n

Even if one ignores all these complexities and replaces real biological neurons with extremely simple simulated ones that are all identical and obey very simple rules.\u00a0<\/p>\n

Currently the most popular model for such an artificial neural network represents the state of each neuron by a single number and the strength of each synapse by a single number.<\/p>\n

In this model, each neuron updates its state at regular time steps by simply averaging together<\/strong> the inputs from all connected neurons.<\/p>\n

Weighting them by the synaptic strengths, optionally adding a constant, and then applying what\u2019s called an activation function<\/strong> to the result to compute its next state.\u00a0<\/p>\n

Activation Functions<\/h3>\n

The easiest way to use a neural network as a function is to make it feedforward<\/strong>, with information flowing only in one direction.<\/p>\n

In case you like math, two popular choices of these activation functions.<\/p>\n