1) You are a workers’ compensation claims analyst working for a large manufacturing organization (worker’s compensation is a worker injury protection; it pays medical and other expenses for workers who were injured on the job). A key part of your responsibilities is to “reserve” a certain amount of funds for each workers’ compensation claim you process, which is necessary for making sure that each injury claim has adequate funds to pay its costs. This requires you to forecast, or estimate the expected future cost of each claim, which is both time-consuming and difficult, as there are multiple factors (such as the nature of injury, worker’s age, the nature of accident, etc.) that determine the ultimate cost of individual claims. Having a lot of historical data (past claims), you decide to use regression analysis to help you forecast the expected cost of the individual workers’ compensation claims.
Discuss how you would use regression to help you with your task—specifically, explain how you would go about evaluating the goodness-of-fit and the predictive efficacy of your model.
Justify your answers with examples and reasoning
2) As a marketing analyst, you are responsible for estimating the level of sales associated with different marketing mix allocation scenarios. You have historical sales data, as well as promotional response data, for each of the elements of the marketing mix. State the differences between the forecasting methods that can be used. Which one would you use and why? If you make any assumptions, state them explicit