Researchers from the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory and the Laboratory for Information and Decision Systems have developed a machine-learning algorithm that can predict the price of Bitcoin.
The researchers collected price data from all major Bitcoin exchanges every second for five months, collecting more than 200 million data points. They then used a technique called "Bayesian regression" to train an algorithm to automatically identify patterns in the data, which were used to predict prices. Every two seconds, the algorithm predicted the average price movement over the following 10 seconds.
"We developed this method of latent-source modeling, which hinges on the notion that things only happen in a few different ways," says principal investigator Devavrat Shah. "Instead of making subjective assumptions about the shape of patterns, we simply take the historical data and plug it into our predictive model to see what emerges."
In the future, the researchers want to expand the scale of the data collection to further hone the effectiveness of the algorithm. The findings indicate predictive powers are often slowed by preconceived notions of what patterns will appear.
From MIT News
View Full Article
Abstracts Copyright © 2014 Information Inc., Bethesda, Maryland, USA
No entries found