#mainTextBox We are now turning the breadboard over to you. Use your own data by changing the input and two desired files (one on L2 and one on the correlation coefficient -- remember to normalize). Once it is trained, turn off the learning (learning check box on the controller inspector) and put probes on the output, desired, and instantaneous cost access points to view the prediction for each point (single step through the data). If you have test data, find out how well the network does on data it has never seen before. Play with the step sizes (open the Inspector on the green "step" icons") to get the fastest convergence without instability. Try adding a learning curve. Explore the system and your new found knowledge!
