There’s a big discussion going on today among the Press, Software Developers, Data and Artificial Intelligence Scientists, Practitioners and Scholars, related to the topic we are covering today.
Many professional use the terms Artificial Intelligence (A.I.) and Machine Learning interchangeably, but in the following articles we are going to break this down for you.
Are Deep Learning, Machine Learning and A.I. the same thing?
Any device that perceives its environment and takes actions in order to maximize its possibilities of success can be said to have some kind of Artificial Intelligence.
More specifically we can say that A.I. exists when a machine has cognitive capabilities, such as problem-solving and learning. Usually associated with a benchmark: the human level, in terms of reasoning, speech and vision, so we can say that AI has three different levels:
- Narrow A.I.: When a machine can perform a specific task much better than a human. This is where most of our technology currently is.
- General A.I.: When a machine can perform any intellectual task as a human would perform.
- Strong A.I.: When a machine can beat humans in several tasks or one complex task.
History tell us about of the first conceptualized A.I. machine, the Perceptron, an algorithm invented in 1957 at the Cornell Aeronautical Laboratory by Frank Rosenblatt. The perceptron is an algorithm for supervised learning of binary classifiers. A binary classifier is a function which can decide whether or not an input, represented by a vector of numbers, belongs to some specific class. It is a type of linear classifier, i.e. a classification algorithm that makes its predictions based on a linear predictor function combining a set of weights with the feature vector. The Perceptron is a single-layer Artificial Neural Network designed for image recognition in the late 50s.
The reason why these algorithms (or machines) are called Neural Networks is because the first practitioners of A.I. thought that this interconnected nodes look like the human neural system. We, humans, have neural networks in our neural system, and these are natural neural networks, while the Perceptron is a rudimentary version of the artificial ones.
As a subset of A.I. we have Machine Learning created in the 1980s. Machine Learning happens when algorithms are trained and learned from past examples in a model that maps features to a corresponding outcome variable. Here relies most of the applications of A.I. for business.
And as a subset of Machine Learning we have Deep Learning, it is called Deep Learning because it makes use of Deep Neural Networks. While Shallow Neural Networks have only one hidden layer between the input and the output, Deep Neural Nets have more than one layer. This field is particularly responsible for all the advancements we had in the recent years in image recognition.
Put simply, deep learning is all about using neural networks with more neurons, layers, and interconnectivity.
Deep learning is one of many approaches to machine learning. Other approaches include decision tree learning, inductive logic programming, clustering, reinforcement learning, and Bayesian networks, among others.
As a summary, Deep Learning, Machine Learning and Artificial Intelligence are not three different things, they are just a subsample of each other.
A.I. refers to devices exhibiting human-like intelligence in some way. There are many techniques for AI, but one subset of that bigger list is machine learning – let the algorithms learn from the data. Finally, deep learning is a subset of machine learning, using many-layered neural networks to solve the hardest (for computers) problems.