attachment_1 (67)

docx

School

Yale University *

*We aren’t endorsed by this school

Course

136b

Subject

Economics

Date

Nov 24, 2024

Type

docx

Pages

2

Uploaded by ogwenogatamu

Report
Part 1:Forecasting IBM’s Earning per Share 1.Apply those models Weighted Moving Average Simple Exponential Smoothing Double Exponential Smoothing (Holt), and Triple Exponential Smoothing models (Holt-Winters). (Include the Additive and Multiplicative versions where the model captures Trend or Seasonality). to the data and choose your best model for forecasting based on 3 indicators. For each model, do the following: a. Split your data into training and testing, where the testing dataset is limited to the last two years of data, and the training dataset covers the remaining years. b. Fit your model using the training dataset and plot the fitted and actual observations in one graph. c. Generate a forecast covering the testing period and plot the fitted and actual observations in one graph. d. Compare the average percentage deviation in the training and testing period. 2.Compare the fit and forecast accuracy of the different models and choose your champion
Part 2:DATA: Consumer Loans: Credit Cards and Other Revolving Plans, All Commercial Banks (CCLACBW027NBOG) Transform the dataset to monthly data and apply a SARIMA(p,d,q)(P,D,Q) model to the data following those steps: 1.Import the data into R from FRED website. 2.After importing it into R, restrict the dataset to the period from April 2014 through end of 2019. 3.Split the data into training and testing set, where the training set covers the training period from the first week in April 2014 until the last week of March 2019, and the testing set covers the testing period from the first week of April 2019 until the end of the year. 4.Train your model on the training set then use the trained model to forecast during the testing period and compare the results of the forecast to the testing set, according to the following steps: a. Identify the model parameters using ACF and PACF. b. Identify the model parameters without having recourse to ACF and PACF. c. Fit the model on the training data using auto.arima function in R. d. Plot the model fit and the actual training set in one graph. e. Use the trained model to create a forecast covering the testing period. f. Plot the model’s forecast and testing dataset inn the same graph. g. Test the model’s forecasting power for 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12 periods ahead, and create a plot that portrays the forecast accuracy for periods 1 through 12.
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