In the world of computers, artificial intelligence is the ability of an electronic device or computer program to use its brain or intelligence to perform different kind of tasks; for example, in the binary world, a computer program may use its artificial intelligence and neural networks to analyze the environment, learn from the environment, and perform a particular task.
As you can see, this is a diagram that an individual can use to understand some of the differences between artificial intelligence and its subfields:
Artificial Intelligence: As previously stated, artificial intelligence (AI) is the ability of an electronic device or computer program to use its brain and intelligence to perform some operations. For example, in software development, when an individual is interacting or chating with a computer program, a computer program can use some conditional statements to evaluate the user input and respond logically to that user, imitating the brain or intelligence of a human being. In the computers world, this imitation is known as artificial intelligence. In addition to that kind of artificial intelligence, there are other kinds of AI, such as machine learning and deep learning. In those subsets of AI, we can say that a computer application is using AI when it is using machine learning features or deep learning features.
Machine Learning: Machine learning (ML) is a subset of artificial intelligence where a computer program can improve itself with time and experience.
Deep Learning: Deep learning (DL) is a subset of machine learning where a computer program can use advanced neural networks to train and improve itself.
In an electronic device or computer application, an artificial neural network (ANN) is similar to the neural network of a human brain, and it is composed by a group of neurons, which will have several inputs and one output. As you can see, those are some of the most important artificial neural networks (ANNs) that an electronic device or a computer application can use to perform its operations:
-Feedforward Neural Networks: In the brain of an electronic device or computer program, when the information is moving directly in one direction among the neurons or nodes of its brain, the agent has a feedforward neural network inside of its brain. To put it another way, an agent will have a feedforward neural network on its brain when the information is moving in one direction among its neurons and without doing circles among nodes.
-Recurrent Neural Networks: The term recurrent neural network (RNN) is derived from another artificial intelligence term that is known as feedforward neural network. In this case, an agent will have a recurrent neural network on its brain when the information is moving in loops among its neurons or nodes to be able to store and remember data from previous experiences.
-Convolutional Neural Networks: A convolutional neural network (CNN) is a deep learning term that is used for those neural networks that are using convolution, which is a mathematical operation where two different kinds of functions can produce a third function. In deep learning, an electronic device or computer program can use its convolutional neural network (CNN) for image recognition, image classification, natural language processing (NLP), and other deep learning tasks.
In the field of machine learning, there is a concept that is called reinforcement learning, also known as RL. What is reinforcement learning? It is an ability of trial and error that an agent is going to use to learn from an unknown environment. While a computer application is learning from an unknown environment, this agent will have the notion of rewards and penalties to be able to learn from the environment and improve its ability to perform a better action in the future.