Geoffrey Hinton, a Canadian scientist, pioneer, and inventor, is known as the “father of machine learning.” He is a genius who helped to develop modern AI. Most of his work is devoted to understanding neural networks and designing machine learning algorithms. Although artificial intelligence and machine learning boomed in 2019, Hinton and other inventors have worked in the field since the 1900s.
Hinton played a crucial role in the history of machine learning, which started back in 1943. His significant role in popularizing backpropagation and neural networks is undeniable. Correspondingly, he authored and co-authored many research papers.
Previously, he worked in favor of backpropagation, the Boltzmann machine, and deep learning. And he is now working on capsule neural networks. Before we talk about Hinton’s contribution to the development of machine learning, there is something you need to know—an overview of Hinton’s journey till now.
Geoffrey Everest Hinton
6 December 1947
University of Cambridge (BA)
University of Edinburgh (PhD)
AAAI Fellow (1990)
Rumelhart Prize (2001)
IJCAI Award for Research Excellence (2005)
IEEE Frank Rosenblatt Award (2014)
James Clerk Maxwell Medal (2016)
BBVA Foundation Frontiers of Knowledge Award (2016)
Nobel prize in computing (2019)
Princess of Asturias Award (2022)
Who is Geoffrey Hinton?
Geoffrey Everest Hinton, popularly known as Geoffrey Hinton, was born on December 6, 1947. He is a British-Canadian cognitive psychologist and computer scientist who pursued his BA from the University of Cambridge and Ph.D. in artificial intelligence. Later worked at famous institutes, including the University of Sussex, the University of San Diego, and Carnegie Mellon University.
Since 2013, Hinton has been dividing his time working with Google and the University of Toronto. Later, in 2017, he co-founded the Vector Institute in Toronto where he works as a chief scientific advisor.
Hinton’s career and research
Before we proceed it’s important to understand the history of machine learning.
Earlier, there was no concept of machine learning. Initially, scientists had only one source of human intelligence, i.e brain. They tried to imitate the functioning of the smallest unit of the brain, neurons and named it an artificial neuron.
However, a single neuron can barely perform any function. Thus, scientists tried to develop a cluster of artificial neurons that can perform a task. A machine performing a human task, that’s the origin of artificial intelligence. The development of artificial intelligence resulted in the emergence of other branches such as machine learning, deep learning, and robotics.
Deep learning is defined as the capability to imitate human intelligence. Many scientists played significant roles in its development. To understand Hinton’s role in machine learning, firstly we have to look at the brief history of deep learning:
Back Propagation – Hinton co-authored the seminal paper “Learning representation by backpropagation errors.” The research paper focuses on backpropagation algorithms. Backpropagation is an algorithm that tests for errors from output to input nodes.
Although there is no direct impact on machine learning, it improves accuracy in data mining, AI, and other similar fields. The key function of backpropagation is to make the system error-free. It improves the efficiency and reliability of the network.
Boltzmann Machine – David Ackley, Terry Sejnowski, and Hinton invented the Boltzmann machine. It’s one of the first neural networks capable of learning internal representation. In other words, it is a neural network with binary nodes. These nodes make decisions based on iteration and the system is trained to detect error signals in each iteration.
The Boltzman machine plays an important role in deep learning. Later after years, they renamed the Boltzmann machine as Deep Boltzmann Machine. This machine had a better structure and understanding than the previous one.
Capsule Networks – Capsule Network is a type of artificial neural network. Hinton proposed the idea of capsule networks back in the 2000s. Hinton and his team attempted to replicate the biological neural network. In simple terms, a capsule network is a system that attempts to perform segmentation and recognition at the same inference.
It took them seven long years to develop a dynamic routing system. The same year, in 2017, Hinton published two open-source research papers on neural networks. He said, “finally, something that works well.” Both papers are available for free.
All these inventions accelerated machine learning.
Hinton’s role in machine learning
Despite being known as the “Father of Machine Learning,” Hinton is best known for his work in deep learning. He worked heavily to promote cognitive science communities in machine learning and deep learning.
Earlier, the idea of imitating human neurons was considered an impossible task. But Hinton always supported the possibility of achieving success in artificial neural networks and machine learning.
Another notable thing he did to create awareness was that he made most of his research papers public. He also rolled out his free course explaining the basics of machine learning. Thousands of students joined the course and gained a real insight into machine learning.
Furthermore, Dr. Hinton also launched his playlist on Neural Network for machine learning. It’s available for free on Youtube.
By this point, you’d already be aware Geoffrey Hinton gained a lot of traction for his significant work. To date, his efforts resulted in popularizing machine learning. Therefore, he is best known as the father of machine learning.