SYS670_FINALEXAM_ROHIT_S_CWID_10466264

docx

School

Stevens Institute Of Technology *

*We aren’t endorsed by this school

Course

EM670

Subject

Industrial Engineering

Date

Dec 6, 2023

Type

docx

Pages

14

Uploaded by PrivateRiverMule27

Report
SYS 670 - Final Exam Fall 2021 Name: Rohit Shegokar CWID - 10466264 Problem 1: 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2002 2005 2008 2011 2014 2017 2020 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 Ice Cream, 1000 gal Frozen Yogurt, 1000 gal The above plotted graph helps in understanding the sales of the given products from the year 1950 – 2020 and thus helps in determining the trend. The sales of ice cream keep on increasing but drops at certain point and the same trend still keeps on. Ice cream, 1000gal model: These are model comparison done for this problem. From these models, I’m considering top 2 for the further analysis. Comparison of top 2 models:
Model 1: Linear (holt) Exponential smoothing
This model is considered as the preferred model because it has a good Rsquare value of 0.945. The Residual graph shows good fit because the values aren't spread out across a large area and the lag value isn't above the blue line. This shows that the curve is well-fitting.
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
Model 2: ARIMA (1,1,1)
 This model has a good R-square value, but it cannot be used for further analysis because of the high probability value. Conclusion: After comparing and analyzing the above models, Linear (Holt) Exponential Smoothing is the best model for the above problem.
From the above model comparison, these are the observations made: Rsquare is low because of the unavailability of full data. Since the dataset is incomplete, seasonal ARIMA can’t be performed. The probability and residuals are not good in any cases. Conclusion: Therefore, none of the models fit for the frozen yogurt. Problem 2: Simple exponential smoothing (zero to one) must be used as it has less error i.e 5.837 and Rsquare is 0.279.
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
Problem 3: The graph shows the monthly average retail price of residential electricity in Arizona from Jan 1990 till Sep 2021. The trend of the graph seems to be varying constantly with same pattern. The retail prices are more or less in the same range for the given years. These are the model comparison I did for this problemThe best model for this problem is Seasonal Exponential Smoothing (12, zero to one). Model Seasonal Exponential Smoothing (12, zero to one) (preferred model):
(a)
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
This model is considered as the best model for this problem because of the following reasons: It has good Rsquare value of 0.939. It also has very good residuals and they are not scattered, they are compressed which is a good fit. The residual graph lag is also good. (b) Calculate the forecast errors and discuss the reasonableness of the forecasts.
The forecasted values are very less reasonable and can be used but not accurate for calculating the data. As can be seen from the errors, they have a wide variation from the true reading. (c) Would it more reasonable to begin the series data at another point in time (later than 1990)? If yes, which year would you recommend starting the time series and why? Yes, 2000. Because at the end of year 1999, the forecast error reaches its second maximum value. From the year 2000, time series would be constant with minimum errors.
Problem 4: The graph shows the monthly computer and electronics manufacturing shipments (in millions of dollars) from Jan 1992 till 2021. The graph's trend appears to be highly complicated, with monthly drops and rises. This isn't a graph with a consistent trend. (a) ARIMA model using the data from 1992 to June 2021 These are the model comparison I did for this above problem. The Rsquraes values are almost similar. Based on the ranking, Seasonal ARIMA(0,0,0)(1,2,1)12 would be a suitable fit for this problem. Seasonal ARIMA(0,0,0)(1,2,1)12 (preferred model):
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
From the detailed analysis of the model Seasonal ARIMA(0,0,0)(1,1,1)12, it has the most suitable Rsquare value and also has a lesser probability with good values of residual curves. There, this is the preferred model. Discuss the presence (or lack) of seasonality in the data . The seasonality occurs because of very close readings occurring in a certain period of time intervals.