John McCarthy, widely recognized as one of Artificial Intelligence (AI) godfathers, defined AI in 1955 as “the science and engineering of making smart machines capable of achieving goals like humans do.” In short, Machines ‘ exhibited human intelligence is Artificial Intelligence.
In 1959, Arthur Samuel defined Machine Learning (ML) as a large AI sub-field dealing with the study field that gives computers the ability to learn without explicit programming. This means that once created, a single program will be able to learn how to do some smart activities outside the programming notion. This contrasts with purpose-built programs whose behaviour is defined by hand-made heuristics that define their behaviour explicitly and statically. So, you might say that Machine Learning is an approach to Artificial Intelligence.
That’s exactly how people learn. We don’t tell them an algorithm/procedure to identify the features when any kid learns to identify objects/person and then decide what it is. We simply show them several examples of that object and then our human brain automatically (subconsciously) identifies the features and learns to identify that object. That’s what a Machine Learning Model actually does.
There is an area often referred to as brain-inspired computation within the fields of machine learning. The human brain is one of the best learning and problem-solving’ machines’ we know. The technique inspired by the brain is actually inspired by the way our human brain works. Our brain’s main computational element is believed to be neuron. The complex connected neuron network forms the basis for all decisions made on the basis of the various collected information. This is exactly what the technique of Artificial Neural Network is doing.
There is an area called Deep Learning (DL) within the neural network domain, where neural networks have more than three layers, i.e. more than one hidden layer. Deep Neural Networks (DNNs) are these neural networks used in deep learning.
Deep learning is, therefore, a technique used to implement machine learning. There are many tasks that machines can do better now than humans thanks to deep learning. One example of this is the classification of images. The winning entry of ImageNet, ResNet, exceeded human-level accuracy in 2015, with a top-5 error rate below 5%. People can classify pictures at an error rate of 5%.
To sum up: Artificial Intelligence is human intelligence displayed by machines, Machine Learning is an approach to achieving artificial intelligence, and Deep Learning is a technique to implement machine learning