Deep learning

 

Deep learning is a class of machine learning algorithms that(pp199–200) uses multiple layers to progressively extract higher-level features from the raw input. For example, in image processing, lower layers may identify edges, while higher layers may identify the concepts relevant to a human such as digits or letters or faces.

Deep learning (also known as deep structured learning) is part of a broader family of machine learning methods based on artificial neural networks with representation learning. Learning can be supervised, semi-supervised or unsupervised.

What is deep learning?

Deep learning is a subset of machine learning in which multilayered neural networks - modeled like the human brain - "learn" from large amounts of data. At each layer of the neural network, deep-learning algorithms perform computations and repeatedly predict, gradually "learn", and gradually improve the accuracy of the result over time.


In the same way that the human brain absorbs and processes information entering the body through the five senses, it swallows deep learning information from multiple data sources and analyzes it in real time.


In-depth learning performs many AI (AI) programs and services that improve automation, and performs analytical and physical tasks without human intervention. Deep learning technology lies behind everyday products and services (such as digital assistants, remote TV with voice playback and credit card fraud detection) as well as emerging technologies (such as self-driving cars).top Ai companies


Feelings of deep learning

The concept of machine learning was first theorized by Alan Turing in the 1950s, but it was not until the mid-1960s that the idea came to fruition when Soviet mathematicians developed the first intermediate set of neural networks.

In the early 1980s, John Hopfield's frequent neural networks made noise, followed by Terry Szynowski's NetTalk program, which could pronounce English words.

In 1986, Professor Carnegie Mellon and computer scientist Jeffrey Hinton - now a Google researcher and long known as the "godfather of deep learning" - were among several researchers who, by showing more than a handful of Learn the use of backpropagation to improve shape and predict word prediction. In 2006, Hinton coined the term "deep learning."


Ian Lekon took the invention of a machine that could read handwritten figures to the next level, followed by other discoveries that were largely under the radar of the wider world. Hinton and Lacon were recently among the three AI pioneers to win the 2019 Turing Award.


Learn how deep work works

Deep learning neural networks (called deep neural networks) are modeled on the way scientists believe the human brain works. They process and reprocess the data, gradually analyzing and modifying the results to correctly identify, classify, and describe the objects within the data.


Deep neural networks are composed of several interconnected layers of nodes, each of which uses more and more sophisticated deep learning algorithms to extract and identify features and patterns in the data. They then calculate the probability or certainty of classifying or identifying an object or information in one or more ways.


The input and output layers of a deep neural network are called visible layers. The input layer is where the deep learning model devours the data for processing, and the output layer is where the final identification, classification, or description is calculated.


In between there are input and output layers of hidden layers, the calculations of each previous layer are weighed and refined by more complex gradual algorithms to zero the final result. This movement of computations through the network is called forward propagation.


Another process called backpropagation detects calculated prediction errors, assigns them weight and bias, and directs them to previous layers to train or modify the model. Together, forward and backward propagation allows the network to predict the identity or class of the object while learning from inconsistencies in the results. The result is a system that works and becomes more efficient and accurate while processing large amounts of data.


The above describes the simplest type of deep neural network in the simplest terms. In practice, deep learning algorithms are extremely complex. And many complex deep learning methods and models have been developed to solve certain types of problems, including the following examples:

Convolution Neural Networks (CNN)

 Used primarily in computer vision applications, it can recognize the features and patterns of a complex image and, ultimately, identify specific objects in the image. In 2015, CNN for the first time improved a human in the object object challenge.


Recurrent neural network (RNN)

Used for deep learning models in which features and patterns change over time. RNNs take data sequences and exit them instead of eating and outputting snapshots. RNNs run emerging applications such as speech recognition and driverless cars.


Deep learning programs

Real-world deep learning programs are part of our daily lives, but for the most part, they are so integrated into products and services that users are unaware of the complex data processing that takes place in the background. Some of these examples include the following:


1. Law enforcement

Deep learning algorithms can identify dangerous patterns that indicate fraudulent or criminal activity by analyzing and learning trading data. Speech recognition, computer vision, and other deep learning programs can improve the efficiency and effectiveness of research analysis by extracting patterns and evidence from audio recordings, videos, documents, which helps law enforcement greatly Analyze data more quickly and accurately.


2. Financial services

Financial institutions regularly use predictive analytics for algorithmic stock trading, assess business risks to approve loans, detect fraud, and help manage credit and investment for customers.


3. Customer service

Many organizations have incorporated deep learning technology into their customer service processes. Bot chats - used in a variety of applications, services and customer service portals - are a simple form of artificial intelligence. Traditional chat bots use natural language and even visual recognition, which is commonly found in call center menus. However, complex Chatbot solutions try to determine multiple answers to ambiguous questions through learning. Based on the responses it receives, the chatbot tries to respond directly to these years or simplify the transmission of the conversation to the human user.


Virtual assistants such as Apple Siri, Amazon Alexa or Google Assistant add a third dimension to the chatbot concept by combining deep learning capabilities and infrastructure technology. These data science innovations enable custom speech recognition and responses, resulting in a personalized experience for users.


4. Health care

Since the digitization of hospital records and images, the healthcare industry has benefited greatly from deep learning capabilities. Image recognition programs can support medical imaging professionals and radiologists, helping them to analyze and evaluate more images in less time.


Special notes

A traditional approach to detecting fraud or money laundering may rely on the amount of the transaction, while a nonlinear deep learning method includes time, location, IP address, retailer type, and any other attributes that may indicate It is fraud. Activity. The first layer of the neural network processes a raw data input such as the transaction value and transmits it to the next layer as output. The second layer processes the information of the previous layer by entering additional information such as the user's IP address and transmits the result.


The next layer captures the information of the second layer and contains raw data such as geographical location and makes the device pattern even better. This continues at all levels of the neural network.

An example of deep learning

An in-depth learning example can be created using the cheat detection system mentioned above with machine learning. If the machine learning system creates a model with parameters built around the number of dollars the user sends or receives, the deep learning method can start based on the results provided by machine learning.

Each layer of the neural network is built on top of the previous layer with added data such as retailer, sender, user, social media event, credit score, IP address and other features that, if processed by humans, may Has been connected for years. Deep learning algorithms are trained to not only create patterns from all transactions, but also to know when a pattern indicates the need for fraudulent research. The final layer transmits a signal to an analyst that may block the user's account until all pending investigations have been completed.

Deep learning is used in all industries for a number of different tasks. Business applications that use image recognition, open source operating systems with consumer-recommended applications, and medical research tools that explore the possibility of reusing drugs for new diseases are just a few examples of deep learning combinations.

Popular posts from this blog

What is crypto wallet

The Best Online store Websites of 2020

Creator of Ltc