BrainCom is an artificial neural network. Utilizing back propagation algorithm it can be used in almost all research fields such as nonlinear regression, forecasting, curve fit, pattern recognition, classification, decision making.

Features: BrainCom is an artificial neural network. Artificial neural networks are analytical systems that address problems whose solutions have not explicitly formulated. Normally the problems to be solved by the classical methods in the computer are problems that can be described explicitly. The program uses back propagation to solve problems in many real-life problems. Non linear regressions, forecasting, curve fitting, pattern recognition, classification, optimization etc are some such examples. The simple interface makes it easy for complete novices to use the system. The interface shows the network stats on the left of the user interaction area. How many inputs are used is indicated. An output specified is also indicated. Number of vectors, topology details (two hidden layers for example) are also available.

Neurons in use, number of iterations to arrive at the solution, total error, % correct, time taken and the date are some of the other details available on this panel. Training parameters can be specified prior to getting the calculations started. These parameters include momentum, rate, parameter for the hidden nodes such as first layer nodes and second layer nodes can be specified here. Other details that can be set are activation function, stopping conditions such as maximum number of iterations, tolerance percentage and correctness level are other things you can specify. The utility can use unlimited number of vectors, inputs/outputs and hidden layer nodes. Training data can be created in spreadsheets and imported into the application as CSV files.

Overall: A nice way of solving some classes of problems.