Neural networks, also known as neural nets or artificial neural networks (ANN), are machine learning algorithms organized in networks that mimic the functioning of neurons in the human brain. Using this biological neuron model, these systems are capable of unsupervised learning from massive datasets.
This is an important enabler for artificial intelligence (AI) applications, which are used across a growing range of tasks including image recognition, natural language processing (NLP), and medical diagnosis. The related field of deep learning also relies on neural networks, typically using a convolutional neural network (CNN) architecture that connects multiple layers of neural networks in order to enable more sophisticated applications.
For example, using deep learning, a facial recognition system can be created without specifying features such as eye and hair color; instead, the program can simply be fed thousands of images of faces and it will learn what to look for to identify different individuals over time, in much the same way that humans learn. Regardless of the end-use application, neural networks are typically created in TensorFlow and/or with Python programming skills.