David Rumelhart

David Rumelhart

Rumelhart was a mathematical psychologist and a student of W. K. Estes. He developed connectionism, an approach to psychology that elucidated areas such as: AI, anthropology, information science, and decision science.

Neural Networks

You are a neural network (see chapter 2). Your senses are connected to your brain and your brain makes decisions for you. Artificial neural networs are common now. Large Language Learning models such as ChapGPT are one example. Both types have much to do with learning as we will see this semester.

In biological networks, learning takes many forms and is widely distributed in the animal kingdom (that include you, btw.) Interestingly, psychologists still do not know all of the mysteries surrounding biological learning. Suffice it to say that the developmental processes that surround growth and maturation include learning. We humans are learning machines.

Speaking of machine learning is more difficult. Artificial networks have three layers: input, hidden, and output. We do not understand what actually happens within the hidden layer of a machine learning network. In machine learning, computers take advantage of their complexity and their ability to simultaneously correlate many instances of a concept. For example, a machine learning device will "understand" what a cat is after "seeing" many thousands of pictures of cats.

A neural network diagram

So, learning itself now has two major facets: biological learning and machine learning.

 


Back to Main Page