How the brain handles big data
"Basically, everything in the brain involves doing calculations on large amounts of data," says neurobiologist and SFI External Professor Charles Stevens of the Salk Institute, the meeting's host.
There are two classes of such calculations: Those hard-wired by evolution and those that have to be learned. Hard-wired calculations, for example, are those involved in making sense of visual data from the eyes.
On the other hand, olfactory data from the nose is more complex than building a picture in three dimensions and can't be hardwired, Stevens explains. There are so many odors and combinations of odors that instead of three visual dimensions to make sense of, olfactory perception has to juggle something like a thousand dimensions to identify a single point.
"That's big data," Stevens says. "So, for example, until a few hundred years ago nobody had smelled coffee. Today everybody can tell the odor of coffee, good from bad coffee, and even whether it's from Starbucks. Evolution could not have wired it in."
The working group's participants – Stevens, Venkatesh Murthy of Harvard, and Stephen Smith of Stanford – also discussed mapping of the brain, cell by cell, which poses a different problem: How do you handle the massive amount of data that would then describe a brain?
The working group took place August 11-24 at SFI.
Provided by Santa Fe Institute