Skip to content

all deep learning basics concepts and the dataset and codes

License

Notifications You must be signed in to change notification settings

gunjanmimo/Deep-Learning-Project

Repository files navigation

deep-learning-a-z

What Is Deep Learning?

Deep learning is an artificial intelligence function that imitates the workings of the human brain in processing data and creating patterns for use in decision making. Deep learning is a subset of machine learning in artificial intelligence (AI) that has networks capable of learning unsupervised from data that is unstructured or unlabeled. Also known as deep neural learning or deep neural network.

How Deep Learning Works

Deep learning has evolved hand-in-hand with the digital era, which has brought about an explosion of data in all forms and from every region of the world. This data, known simply as big data, is drawn from sources like social media, internet search engines, e-commerce platforms, and online cinemas, among others. This enormous amount of data is readily accessible and can be shared through fintech applications like cloud computing.

However, the data, which normally is unstructured, is so vast that it could take decades for humans to comprehend it and extract relevant information. Companies realize the incredible potential that can result from unraveling this wealth of information and are increasingly adapting to AI systems for automated support.

Deep Learning Versus Machine Learning

One of the most common AI techniques used for processing big data is machine learning, a self-adaptive algorithm that gets increasingly better analysis and patterns with experience or with newly added data.

If a digital payments company wanted to detect the occurrence or potential for fraud in its system, it could employ machine learning tools for this purpose. The computational algorithm built into a computer model will process all transactions happening on the digital platform, find patterns in the data set and point out any anomaly detected by the pattern.

Deep learning, a subset of machine learning, utilizes a hierarchical level of artificial neural networks to carry out the process of machine learning. The artificial neural networks are built like the human brain, with neuron nodes connected together like a web. While traditional programs build analysis with data in a linear way, the hierarchical function of deep learning systems enables machines to process data with a nonlinear approach.

A traditional approach to detecting fraud or money laundering might rely on the amount of transaction that ensues, while a deep learning nonlinear technique would include time, geographic location, IP address, type of retailer and any other feature that is likely to point to fraudulent activity. The first layer of the neural network processes a raw data input like the amount of the transaction and passes it on to the next layer as output. The second layer processes the previous layer’s information by including additional information like the user's IP address and passes on its result.

The next layer takes the second layer’s information and includes raw data like geographic location and makes the machine’s pattern even better. This continues across all levels of the neuron network.

Applications of Deep Learning

To better understand what Caffe2 is and how you can use it, we have provided a few examples of machine learning and deep learning in practice today.

computer vision shades icon Computer Vision

Computer vision has been around for many years and has enabled advanced robotics, streamlined manufacturing, better medical devices, etc. There is even license plate recognition to automate giving people tickets for a number of moving violations like speeding and running red lights. Neural networks have significantly improved computer vision applications. Photo processing is being used for object recognition, answering questions such as “Is that a cat or a dog?”. Video processing is being used to automate scene classification or people recognition, answering questions such as “Is that a helicopter? Is there a person in the helicopter? Who is that person?”.

speech recognition waveform icon Speech Recognition

Many readers may have been exposed to Apple’s Siri. This digital assistant’s core interaction with users is through voice recognition. You ask Siri for directions, to make appointments on your calendar, and to look up information. Its ability to understand a variety of accents in English, let alone its multilingual settings and capabilities, is based on many of the enhancements made in Siri since 2014. These enhancements were accomplished through the utilization of Deep Neural Networks (DNN), Convolutional Neural Networks, and other advances in machine learning.

Translation

Another useful application of neural networks is with translation between languages. Translations can occur via voice, text, or even handwriting. One of the Caffe2 tutorials shows how you can create a basic neural network that can identify handwriting of English text with over 95% accuracy. It is not only highly accurate, it is extremely fast.

Chat Bots

There are currently useful interaction with simple AI’s. A common simple AI is a chat bot.

A chat bot could be in action when you click on the support link on your bank’s website or favorite shopping website. The “how may I help you?” response can be a fully automated program that reads your text and looks for related responses, or, in the most simplest form, can redirect you to an appropriate live agent. As more complex bots are written using DNN, their ability to understand your statements, and more importantly, the context, the bots will be able to hold longer, more meaningful conversations without you even realizing you are not chatting with a real person.

IoT

As we explore the full impact and capabilities of the Internet of Things (IoT), where common technology communicates with you - from your fridge, to your security system, to individual lights - a fairly simple AI can automatically (and in real-time) review security camera footage, faceprint visitors to distinguish between homeowner, guest, and trespasser, and adjust lighting, music, and alarm sounds accordingly. How the system distinguishes between parties can be accomplished by training a DNN and then a variety of systems such as AWS’s IoT platform can wrap this core detector to provide responses and actions.

Medical

Customs agencies have used thermal image processing to identify people who may be suffering from a fever in order to enforce quarantines and limit the spread of infection disease. Image segmentation is a common task for in medical imaging to help identify different types of tissue, scan for anomalies, and provide assistance to physicians analyzing imagery in a variety of disciplines such as radiology and oncology. Medical records can be processed with ML and DNN to find insights and correlations in these massive data sets.

Other applications

Deep learning and neural networks can be applied to ANY problem. It excels at handling large data sets, facilitating automation, image processing, and statistical and mathematical operations, just to name a few areas. It can be applied to any kind of operation and can help find opportunities, solutions, and insights. Depending on your role you may find a different attractor for Caffe2 and deep learning.

Business person - how can it make my company money, save costs, increase margins, find new markets or opportunities

Marketing person - find new markets, target within markets, increase effectiveness of marketing, personalization

Product person - enhance products or even create new products with AI and NN at its core

Data person - analyze massive quantities of data to find trends and predictors, and develop new models for any industry

Developers & engineers - ultimately there will be demand from so many industry sectors to utilize deep learning that incorporating it into platforms will be required even if you’re not involved with creating, researching, or refining the deep learning systems themselves

Academics - refinement of existing models, creation of new models, algorithm development, and more intelligent neural networks are forthcoming and there’s a wide open arena of opportunities for academics to help progress DNN and AI.