Directed Research Project Details
Problem Solving Environments and the Neural Network Driver
| Author: | Joel Berghoff |
|---|---|
| Date: | December 07, 2005 |
| Time: | 9:00 |
| Location: | 220 Deschutes |
| Committee: | John Conery (Chair) Shawn Lockery (Biology Dept) Dejing Dou |
Abstract
Problem Solving Environments help scientists perform their work by providing tools to facilitate research. Using environments that facilitate training or simulate the operation of artificial neural networks (ANNs) usually requires the download and installation of software that is often platform-dependent, programming language-dependent, cumbersome to install, or all of the above. The Neural Network Driver is a browser-based tool for training (and eventually simulation of) ANNs that allows users to store experimental parameters and results in a database. This information can later be recalled and used to provide the starting point for additional training or simulation. The browser-based nature of the application obviates the need for downloading additional software while the database backend allows users to store and access data remotely. These same features also make collaboration straightforward.