6. Identify a problem that would be well suited for a neural network. Explain how the system would be trained (e.g., what existing data can be used?). Explain why you think the problem needs a neural network and what benefits can be gained.
Answer is variable. The main issues are to find a problem that requires pattern matching. The common ones of OCR, speech recognition, facial recognition and so on should be avoided or downgraded since they are described in the text. More interesting applications include systems for prediction, both in business and other areas such as disease prevention. One challenge is finding an adequate number of accurate test data that can be used to train the system. If you train it with weak data, the system will not be very reliable.