California Scientific  *  BrainMaker Neural Network Software  *   800-284-8112 * 916-372-6800
California Scientific      BrainMaker Neural Network Software
Home Applications What Are Neural Networks Products Ordering Info Tech Support

Currency Predictions
Corporate Bonds
S&P 500 Index
Stock Price Prediction
Stock Prophet
Natural Gas Prices

Direct Mail
Credit Scoring
Real Estate Appraisal
Bad Cops
Finding Gold
Jury Summoning
Highway Maintenance

Breast Cancer
Hospital Expenses
Functional Recovery
Heart Attack Diagnosis
ER Test ordering
Psychiactric Care
Giant Cell Arteritis

Thoroughbreds 1
Dog Racing

Solar Flares
Protein Sequencing
Mosquito Identification
Weather Forecasting
Predicting El Nino
Air Quality

Welding Quality
Plastics Testing
IC Quality Control
Paper Quality
Concrete Testing
Beer Quality
Enzyme Synthesis
Steam Quality

Pattern Recog.
Speech Recognition
Article Classification
Working with Chaos
Chemical Drawings
Sequence Prediction

BrainMaker Predicts Detrimental Solar Effects

Dr. Henrik Lundstedt of Lund Observatory, Sweden, has trained neural networks to predict solar-terrestrial effects such as disturbances in the earth's magnetic fields. The disturbances have been known to cause blackouts, power plant shutdowns, corrosion in pipelines, disruptions in radio and television transmissions, malfunction of geological survey equipment, satellite tracking problems, and other detrimental effects. Being able to predict these occurrences helps prevent disasters.

The major cause of disturbances on earth are certain behaviors of the sun's solar wind. The solar wind is caused by several things such as coronal mass ejections or CMEs (which can trigger flares), and coronal holes. The neural network inputs consist of 37 known values of solar-terrestrial phenomena such as coronal mass ejections, coronal holes, solar sector boundaries, and proton events. The values are input as changes over the last four days. There are eight output neurons. The first output represents whether geomagnetic activity is expected to be quiet for the next day. The second, third and fourth outputs represent whether the activity is expected to be of a minor, major, or severe storm character. The fifth through eighth outputs predict the same items two days ahead.

The neural networks were trained with seven months of solar data from various US databases from CSSA, Stanford, CA; NOAA/SEL, Boulder, CO; and SacPeak/AFGL, NM. A period of data from June 6 - 21, 1990 was omitted from the training data and used for testing. During that period three major storms and one minor storm occurred. However, the traditional prediction method (NOAA/SEL) predicted no major storms and one minor storm which was in fact one of the major storms. The neural network did much better. It predicted two of the three major storms and the minor storm, and predicted a minor storm for the third major storm.

coronal mass ejections
coronal holes
solar sector boundaries
proton events

quiet tomorrow
minor storm tomorrow
major storm tomorrow
severe storm tomorrow
quiet in two days
minor storm in two days
major storm in two days
severe storm in two days

Home Applications What Are Neural Networks Products Ordering Info Tech Support
California Scientific      BrainMaker Neural Network Software

California Scientific • 800-284-8112 • 916-372-6800