1- A client asks your company to develop a system for a classification problem (e.g., medical insurance fraud detection). Because of the user needs, he requires the final developed system to be able to provide explanations about the system result (i.e., how the system result is generated). The design team proposes a BP neural network system for this application. Do you agree that it will be an effective approach to the problem? Please explain your thoughts .
2-In the class, we introduced several factors that may affect the group process and outcome, such as group size and status effect. How about the levels of familiarity among group members in traditional face-to-face groups without GSS support? In other words, do you think different levels of familiarity among members in a group can affect the group process and outcome? Please explain why or why not
3- Insurance claims adjusters are under great pressure to process insurance claims quickly. Now claims adjusters must determine liability, estimate damages, and deal with false or padded claims in a cost-efficient and timely manner. Extra investigation of a suspicious claim adds expense and delays settlement, but inaction could result in payment of a fraudulent claim and the potential for inviting future bogus claims. It is a difficult dilemma for even the most seasoned claims adjusters as most claims have some suspicious elements. Now you are asked to propose a DSS solution to help the claim adjusters identify and red flag various kinds of fraud. Could you please choose and justify which of the decision technology/approaches, expert systems or neural networks, that should be used and how your solution would work ?
4- A financial company hires your team to develop back-propagation neural network(s) for predicting the next-week trend of five stocks (i.e. go up, go down, or remain the same). In the meantime, the company also provides you the data for each stock in the past 15 years. Each data record consists of 20 attributes (such as index values, revenues, earnings per share, capital investment, and so on). Please answer the following questions:
1) The team member A suggests that you should develop a single neural network that can handle all these stocks. But member B insists that you have to develop five networks (one for each stock). Whom do you think is correct and why?
2) During training, at a certain point, you notice that the error rate (of each training cycle) has been oscillating (i.e. it decreases in the round n, increases in the round n+1, and decreases again in the round n+2, and so on). What might be the reason for this phenomenon and what you should do about it?