Deep Network
In a previous post, the process of how an ML is trained was examined. In that post two specific terminologies were mentioned, input and output. Inputs are the signals or data received by the system while outputs are the signals or data sent from it. Both are major components to our next topic, understanding the functions of deep learning.
Deep what?
Deep learning is known as “a branch of machine learning” that is based off of ANN (Artificial Neural Network) which contains units (neutral networks) that are laid out in a series of layers that form the entire ANN system. A single layer can range from a dozen to millions of units, it all depends on how complex neural networks are required to learn hidden patterns in a dataset. Normally within every ANN there is an input layer, an output layer and hidden layers within those input nad output layers.
Units are interconnected from layer to layer. Such connections hold a strong wight in determining how much units influence each other. Neural networks data learning increases whenever units transfer data to each other. Neural networks learn from outside data (external data, third-party data, or public web data) received from input layers, once it’s receives an already analyzed data, it is then converted into data that is valuable enough to passs through the output layer. At the end, the output layer provides an output in the form of a response to the provided input data.
To put it simply an input layer is where the deep learning model ingests the data for processing, and the output layer is where the final prediction or classification is made. The model’s performance depends on how data was received and processed from layer to layer.
Applications
Deep learning is reportedly able to learn features from data, which is why it has been used to complete tasks such as image recognition, speech recognition, object detection, image segmentation. language translation, and natural language processing.
Wondering how deep learning is used in the real world, well look no further. Neural networks are heavily used in:
Social media: Certain platforms send out ‘People you may know’ suggestion notifications, it basically alerts people of possible real life connections they have to people found on that platform as a way to have them reconnect thus resulting in having both parties follow each other.
Although it may look like a magical or sometimes creepy feature, it is achieved with the help of Artificial Neural Networks that analyzes a persons profile, interests,and list of followers to calculate the possible connection between people registered on that platform