Deep learning explained by example

Original Source Here

Deep learning explained by example

Deep learning is an advanced branch of machine learning that enables computers to solve complex problems from driving a car to successfully flying a helicopter without strictly learning from simulations. The two main branches of deep learning are computer vision and natural language processing.
Deep learning is a combination of mathematics and Neurobiology, the artificial intelligence science aims to project the human learning process on computers so that computers could learn and improve from experience. As mentioned earlier artificial intelligence science is based on how we learn, deep learning is accomplished by imitating human brain architecture. Developing networks of connected processing units called “neurons”. Our brains are composed of 86 billion neurons connected by neurotransmitters. Our senses provide input to the brain, in the brain the information is transmitted from one neuron to another each neuron may add or even alter the information until reaching the final node in which a decision will be taken. think of a child’s first encounter with fire. The child is fascinated by that light fixing his eyes on the fire and could hear the fire voice, as the child reaches to the fire, he feels its warms until he touches it leading to burning. When the child a signal is transmitted from the child’s eyes, ears, and hands to the brain, that signals reach the brain and begin their journey in the nodes, after reaching the final node the brain concludes to identify this new object by its characteristics illuminating, heat, and voice. And that fire is harmful. For simplicity we assumed that the child only learns the object characteristics after pain and this is not true in real life learning process is always on and each aspect of the object will be learned in the precise second of the encounter. one input may carry more important information, so each connection will have a degree of importance in the decision-making process. In the child’s case, the most input was the pain upon touching, in this scenario the brain will increase the weight of the nodes carrying that information. Deep learning is composed of a neural network that has input, processing “hidden” layers, and an output layer. A deep learning network must have more than one processing layer. Each deep learning network is designed and trained to perform a task. In the training process, the degree of importance of each node will be calculated and a corresponding weight will be assigned to that node.


Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot

%d bloggers like this: