ISM 5219 Test 3

You may work in groups of 1, 2, or 3 class members. Use Pattern Recognition Workbench (PRW) and neural networks.

Download data for prime rate and CPI, at least 30 years of concurrent monthly data for each. Divide this data into 2 approximately equal intervals (a learning sample and a test sample). Calculate change in each series for 3 month intervals.

(a) Fit an autoregressive model to change in prime rate in the earlier interval (learning sample) for each time series. Test that model on the out-of-sample interval (the test sample). Evaluate the models in-sample and out-of-sample. Present your results in tabular form with a narrative explanation.

(b) For the intervals of prime rate and CPI data, fit a model in which change in prime rate is the dependent variable and some lagged period change in prime rate is one independent variable and some lagged period change in CPI is the other independent variable. Fit your model in-sample, on the earlier interval, and evaluate it in-sample, on the earlier interval,  and out-of-sample, on the later interval. Present your results in tabular form with a narrative explanation.

(c) Compare the forecasts (a) and (b) on out-of-sample performance. Use "class example 06 16.xls" as a guide.

Put all of this work into a .DOC file and attach it to an email with the subject "ism 5219 test 3 your name" and send the email to william.leigh@bus.ucf.edu by 10 pm on the night of the test (6/30/2005). If there are multiple members of your group have as the subject "ism 5219 test 3 group", and list the group member names in the body of the email.

Thank you.