WELCOME

The Open Water Neural Imaging Project is, unfortunately, one of

Oct. 16, 2018 by Carolyn Mueller

The Open Water Neural Imaging Project is, unfortunately, one of the few open science projects we know of. The project is led by a group led by former MIT professor of neural science, neuroscientist, and neurosurgeon, Jonathan Bausch, and funded by the NSF, the National Science Foundation, and the National Institutes of Health. Bausch is a proponent of "deep learning" as he explains:

As it turns out, neural networks can be used to create new kinds of content without any computational effort. Instead of just searching for patterns in a stream of neurons, neural networks can be used to learn new ways to learn, and so on. Nowhere is this more evident than deep neural networks (which are not only used to learn new ways to learn, but also to learn different ways to learn) using a variety of different types of processing.

The Neural Networks project has a great deal in common with the Open Water work. The goal is to create a "deep neural network" that can learn a set of neural networks, without any computational effort.

In addition, to be able to learn how to build new neural networks, the neural network must first establish how many neurons it can build. As Gazzaley and I note, even though neural networks can learn to build new neural networks—as long as they don't use any computational resources—they have yet to show how to do so.

The goal of neural networks is to learn new ways to learn. So, to build new neural networks, researchers must either establish what neural networks can learn to build, or design new neural networks.

In fact, there are still more ways to build neural network learning programs, and the goal isn't just to learn new ways to learn new neural networks; the goal is to design a neural network that can learn new kinds of content without any computational effort.