Wolfram Computation Meets Knowledge

Wolfram Summer School


Garrett White

Summer School

Class of 2008


At age 15 Garrett White had his first psychology class, and his eyes were opened to a beautiful experience. For the first time he saw himself more clearly through the framework of a larger perspective. He was hooked. After studying psychology further he failed to find what he expected. Still searching at the university he found the sciences to hold beautifully coherent bodies of thought but none were really related to him as a person. Finally, settling in the field of neuroscience, he feels at home.

Currently, he enjoys meeting new people and traveling but also spending time in the forests of Oregon. His favorite hike was a 6-day, 120-mile trek along the Pacific Crest National Scenic Trail carrying only 20 pounds, including food, water, stove, shelter, clothes, and first aid. He found that minimalist experience personally meaningful in that he almost felt that he understood how he was so well suited to excel in such an environment as compared to any other living thing in the forest. This is an example of seeing himself in a new way within a larger context.

Searching for new experiences along these lines is what motivates him and makes him who he is. Oh, and as for that degree in psychology, he did find that he is his best when around those who challenge him in a supportive way. He likes to try to learn failures as much as from successes.

Project: EEG Analysis and NKS Systems

White is currently employed as a data analyst for encephalographic (i.e. brain wave) data of epileptic patients. Several electroencephalographic (EEG) signal structures are diagnostic for this population. These are spike and seizure waveforms. Sleep spindles, wave packets of 1-2 seconds with a football shape, are common in the population at large. Both cellular automata and brain waves exhibit the characteristics of a chaotic system.

In order to investigate whether further similarities exist, the NKS systems must be considered as 2D waves or “NKS signals” similar in appearance to EEG waves. One way this can be accomplished is by a Gaussian convolution, performed by generating a matrix with the values representing a 2D Gaussian distribution. The next step is to impose this matrix onto an NKS system through simple component-by-component multiplication. By taking the total average values of the imposed matrix, a single data point is provided. Incrementally moving this matrix through the NKS system provides a series of data points resembling an EEG. Further searching of different NKS systems may expose the specific waveforms mentioned above.

If NKS systems generate similar waveforms as EEG signals then the next step would be to investigate the analytical contents of these NKS signals. Investigation with statistical tools such as independent component analysis and principal component analysis as well as more traditional Fourier transformations of the analytical contents may be explored. Certainly if the raw form of EEG and NKS signals look similar in form then their analytical components will produce some similarities.

The question then becomes what similarities are in common, and whether these commonalities can lead to a method of investigation for EEG signals. Essentially, this is a search for a novel EEG data analysis technique using the tools of the NKS methodology.

Favorite Radius 3/2 Rule

Rule 22904