Question

Can someone please answer the following 10 Multiple choice questions? Thank you.

1. An analyst needs to build a forecasting model but he is not sure whether it has trend or seasonality components. Which of the following can help the analyst build decide on the forecasting model. Check all that apply.

		Assume it has only seasonality but not trend, because time series with trend are rare.
		Try different models and pick the best one.
		Try plotting the time series and visually determine what components it has.
		Assume it has only trend but not seasonality, because time series with seasonality are rare.

2. _____ distribution is a _____  probability distribution whose mean equals its variance.

			Exponential, distrete
			Poisson, continuous
			Exponential, continuous
			Poisson, discrete

3 . After building a linear regression model, an analyst examined the residuals, which seemed to not be uniformly spread around the regression line, but rather seemed to spread around a curve. Which of the following, the should analyst consider?

		Using nonlinear programming
		Using nonlinear regression analysis
		Using Monte Carlo simulation analysis
		Removing one or more independent variables

4. An alternative medicine researcher is studying whether a newly developed herbal extract lowers blood pressure. Which one of the following confidence interval analyses can help the researcher in her study?

		Difference between proportions
		Difference between means, independent samples, unequal variances
		Difference between means, paired samples
		Difference between means, independent samples, equal variances

5. Poisson distribution can be used to model _____.

		Number of patients arriving at a clinic in a given period of time
		The outcome of flipping a coin
		Blood sugar level in a large population
		Passenger arrival time minus scheduled departure time

6. Assume you are working on a forecasting project, where you are expecting extreme outliers. Which one of the following forecast error measures should you consider? Why?

		MAD because it penalizes outliers disproportionally more (quadratically), hence it is suitable for problems with outliers.
		MSE because it penalizes all the deviations consistently (linearly), hence not sensitive to outliers.
		MAD because it penalizes all the deviations consistently (linearly), hence not sensitive to outliers.
		MSE because it penalizes outliers disproportionally (quadratically), hence it is more suitable for problems with outliers.

7. A group of researchers are studying the difference in the highway gas mileage of the new releases of 2 different car models from 2 different manufacturers, namely Toyota Corolla and Honda Civic. Which one of the following confidence interval analyses can help the researchers in their study?

		Difference between means, independent samples, equal variances
		Difference between means, independent samples, unequal variances
		Difference between means, paired samples
		Difference between proportions

8. An analyst is studying the relationship between the overall productivity of a manufacturing facility and the length of the workers' shift. The analyst decided to apply linear regression analysis to a data set from 91 different manufacturing facilities that belong to the same manufacturer and produce the same type of product. The facilities have different shift lengths, which are between 4 hours and 11 hours. The regression statistics and ANOVA returned acceptable values. Which of the following helps the analyst the most in deciding the relationship between productivity and shift length?

		Rate regression indicator
		Residuals
		Coefficient
		Intercept

9. Assume you are working on an optimization problem where the fraction of sulfur in a new chemical product has to be constrained so it does not exceed 10%. There are 2 ingredient of the product and both of them contain sulfur. The amounts used from these 2 ingredients are denoted X and Y. These 2 ingredients contain 12% and 5% sulfur, respectively. Which of the following correctly expresses the constraint? Check all that apply.

		(0.12X + 0.05Y) / (X + Y) <= 0.1
		0.02X - 0.05Y <= 0
		0.12X + 0.05Y <= 0.1
		0.02X - 0.05Y + 0.1 = 0

10 An R&D engineer is in the process of building a forecasting model for a business problems. He consulted an expert from the operations department of his company. The expert expressed that the business logic dictates that the data series used in the forecasting model is highly expected to incorporate correlations between consecutive values for 2 lags. Which of the following models should the engineer choose?

		Linear optimization model
		Causal model
		First-order autoregressive model
		Second-order autoregressive model

Answer 1