Question 1

  1. R2is 0.8112 which means that the model can explain 81% variance in the price.
  1. Null Hypothesis, H0 = B1 = B2= B3 = B4 = B5 = 0, which means that none of the independent variables has any predictive power

Alternative Hypothesis, Ha= B1 = B2= B3 = B4 = B5 != 0, which means that all of the independent variables havepredictive/explanatory power

Since the F-statistic is significant, we can reject the null hypothesis and agree that all 5 variables jointly explain the variability in price.

  1. Null Hypothesis, H0 = Bvariable= 0, states that the independent variable doesn’t have any significant predictive power

Alternative Hypothesis, Ha= Bvariable != 0, states that the independent variable has significant predictive/explanatory power

Since the p-value of CVN, DPC and GDPN is less than 0.05, they are significant and we can reject the null hypothesis and agree that these 3 variables individually explain the variability in price.

On the other hand, IPC and PP have p-values greater than 0.05, therefore they are not significant predictors. We cannot reject the null hypothesis.

  1. Schutt and VanBergeijk should have concluded that there exists international price discrimination because the overall model is significant as the F statistic is highly significant at 1% significance level.

Question 2

  1. Using corner point method, the maximum feasible solution is $560. It is calculated by plugging in values of points A, B and C in the cost function 5x + 6y.

A (0,60): 5(0) + 6(60) = $360

B (100,10): 5(100) + 6(10) = $560

C (105,0): 5(105) + 0 = $525

Out of these, maximum solution is $560 at x =100 and y = 10.

Question 3

  1. Optimal solution is shown in yellow for the quantities of each product and in green for minimum cost.

  1. Minimum cost per meal is $1.75
  2. Yes, the solution is sensitive to changing food price. The price of milk when reduced to $0.14 per pound would make a change to the optimal solution.

Question 4

  1. 4-period Moving Average

Exponential Smoothing Forecast

Linear Trend Forecast

  1. Moving Average:

MAD or Mean Absolute Error is 5339036.62

MSE (Mean Square Error) = (RMSE)2 = 36357174748679.9

Exponential Smoothing:

MAD or MAE = 8227972.035

MSE = (RMSE)2 = 85105506704290.40

Linear Trend Forecast:

MAD = 4256552.5

MSE = 2.68409E+13

(Calculations are shown below for Linear Trend Forecast)

  1. Plots are shown below:
  1. The forecasts for 4 quarters of 2014 are shown above in the snapshots. Since the Root Mean Square Error of Linear Trend model is the least out of the 3, this is the most effective forecast.
  2. As the data is time series in nature, I may want to try ARIMA forecasting because it not only captures moving averages but also auto regression in the historical values and the essence of stationarity in the data.

Question 6: