{"id":21207,"date":"2018-10-07T08:30:06","date_gmt":"2018-10-07T15:30:06","guid":{"rendered":"https:\/\/insidebigdata.com\/?p=21207"},"modified":"2018-10-08T08:51:00","modified_gmt":"2018-10-08T15:51:00","slug":"introduction-deep-learning-neural-networks","status":"publish","type":"post","link":"https:\/\/insidebigdata.com\/2018\/10\/07\/introduction-deep-learning-neural-networks\/","title":{"rendered":"An Introduction to Deep Learning and Neural Networks"},"content":{"rendered":"

It seems as if not a week goes by in which the artificial intelligence concepts of deep learning and neural networks make it into media headlines, either due to an exciting new use case or in an opinion piece speculating whether such rapid advances in AI will eventually replace the majority of human labor. Deep learning has improved speech recognition, genomic sequencing, and visual objection recognition, among many other areas.<\/p>\n

The availability of exceptionally powerful computer systems at a reasonable cost, combined with the influx of large swathes of data that define the so-called Age of Big Data<\/a><\/u> and the talents of data scientists, have together provided the foundation for the accelerated growth and use of deep learning and neural networks.<\/p>\n

Companies are now beginning to adopt AI frameworks and libraries, such as MxNet<\/a><\/u>, which is a deep learning framework that gives users the ability to train deep learning models using a variety of languages. There are also dedicated AI platforms aimed at supporting data scientists in deep learning modeling and training which professionals can integrate into their workflows.<\/p>\n

It\u2019s important, though, to specify that deep learning, neural networks, and machine learning are not interchangeable terms. This article helps to clarify the definitions for you with an introduction to deep learning and neural networks.<\/p>\n

Deep Learning and Neural Networks Defined<\/strong><\/p>\n

Neural Network<\/u><\/p>\n

An artificial neural network, shortened to neural network<\/em> for simplicity, is a computer system that has the ability to learn how to perform tasks without any task-specific programming. For example, a simple neural network might learn how to recognize images that contain elephants using data alone.<\/p>\n

The term neural network<\/em> comes from the inspiration behind the architectural design of these systems, which was to mimic the basic structure of a biological brain\u2019s own neural network so that computers could perform specific tasks.<\/p>\n

The neural network has a layered design, with an input layer, an output layer, and one or more hidden layer between them. Mathematical functions\u2014termed neurons\u2014operate at all layers. Neurons essentially receive inputs and produce an output. Initially, random weights are associated with inputs, making the output of each neuron random. However, by using an algorithm that feeds errors back through the network, the system adapts the weights at each neuron and becomes better at producing an accurate output.<\/p>\n

\"\"

Image credit: Shutterstock<\/p><\/div>\n

The learning occurs in a neural network by feeding it labeled input and output data, and the network improves its performance by feeding it more and more data. This form of learning is supervised learning<\/a><\/u> because it requires data scientists to provide the algorithm with labeled data for the learning to occur.<\/p>\n

Deep Learning<\/u><\/p>\n

Deep learning is a method of learning that occurs when computer systems learn how to recognize patterns and classify things using raw, unlabeled data (unsupervised learning) instead of the task-specific algorithms of standard neural networks that rely solely on supervised learning.<\/p>\n

Deep learning models can perform both unsupervised and supervised learning. The characteristic feature of the neural networks underlying these models is multiple hidden layers, hence the term \u201cdeep learning\u201d.<\/p>\n

Because real-world data such as images, video, and IoT sensor data are typically unstructured and not convenient to process, deep learning models meet the need of getting computers to learn autonomously by discovering characteristic features of data using special algorithms.<\/p>\n

The aspects of modern technology that facilitate deep learning include the availability of large-scale datasets, cheaper access to fast graphics processing units (GPU)<\/a><\/u> that can rapidly perform mathematical calculations, and improved training algorithms.<\/p>\n

Deep learning, machine learning, and AI<\/strong><\/p>\n

Putting it together in a broader context:<\/p>\n