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

What are NN's Designing NN's MMX Computers References Reviews of BrainMaker Why BM is best Products BrainMaker BrainMaker Professional Competitor Genetic Training Intro to NN Book Do I need Professional? NN Books For Sale Ordering Info International Shipping Order Form Price List Upgrade Order Form Tech Support Running Competitor Financial Networks Is my data OK? Using hidden neurons Large datasets BrainMaker's precision Predicting Horses Setting Min/Max Year 2000 Article Reprints Testing Concrete Decoding Sequences Hospital Costs Downloads MMX test software order form upgrade form Pro Forma invoice Applications Stocks Currency Predictions Corporate Bonds S&P 500 Index Stock Price Prediction Stock Prophet Natural Gas Prices Business Direct Mail Credit Scoring Real Estate Appraisal Bad Cops Finding Gold Jury Summoning Highway Maintenance Medical Breast Cancer Hospital Expenses Functional Recovery Heart Attack Diagnosis ER Test ordering Psychiactric Care Giant Cell Arteritis Sports Thoroughbreds Thoroughbreds 1 Dog Racing Science Solar Flares Protein Sequencing Mosquito Identification Spectroscopy Weather Forecasting Predicting El Nino Air Quality Manufacturing 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

What's the right number of hidden neurons? When you're training a network with BrainMaker, the number of hidden neurons is critical. Too few, and there's not enough available "brain" to learn your problem. Too many, and the network "memorizes" instead of "learns". Here's how to tell: Turn on Display Histograms- we leave this on essentially always. If, at the end of training, your histograms are still bell curve shaped, you can almost certainly reduce the number of hidden neurons, which will very likely improve your network's predictive powers.   If your histograms are relatively flat, you're probable very close to the optimum number of hidden neurons. If your histograms are bunched up at the left and/or right side of the graph, with little near the middle, your network is probably already brain-dead, and will never learn. You need to add hidden neurons, or investigate your data more closely. The most important thing at first is to get the network to train at all. If you're having trouble, try liberally adding hidden neurons, up to about three times as many hidden neurons as you have inputs. After you've successfully trained, then you can try again while cutting back on the number of neurons, thereby trying to build a network with improved predictive powers. If this doesn't work, you probably have problems with your data. Kolmogorov, a Russian mathematician, showed in the 1950's that the correct number of hidden layers is always either one or two. In our experience, one is the correct number about 85% of the time, and two about 15%. Using more than two layers of hidden neurons is a complete waste of time. We allow up to eight hidden layers. It's also OK with us if you toss your PC into a swimming pool, but this will not improve it's performance.

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