Sample For Final Examination STA 3155
Note: Support your answers by calculatios/reasons
Q1. Encircle T (for True) and F (for False) in the following statements:
i. Exponential smoothing weighs heavily remote time series values (T / F)
ii. Moving averages
Chapter 4
Multiple Linear Regression
4.1. The Linear Regression Model:
General Linear Regression Model:
y = 0 + 1x1 + 2x2 + 3x3 + + kxk +
4.2. The Least Squares Estimates, and Point Estimation and Prediction
A point prediction of an observed value of dep
Christina Hanchi
STA 3155 EMWA
April 20th, 2015
SAS Assignment
Logarithmic Model
Data from Chapter 6 page 322
This is a logarithmic regression model. In the first plot of y*time, the data is not linear. By using
logarithms, the data can be transformed in
SAS Output
Page 1 of 4
file:/C:/Users/YSabban/AppData/Local/Temp/SAS%20Temporary%20Files/_TD4232_B. 3/21/2016
SAS Output
Page 2 of 4
file:/C:/Users/YSabban/AppData/Local/Temp/SAS%20Temporary%20Files/_TD4232_B. 3/21/2016
SAS Output
Page 3 of 4
file:/C:/Use
SAS Output
Page 1 of 11
The SAS System
The CORR Procedure
5 Variables: y x1 x2 x3 x4
Simple Statistics
Variable
N
Mean Std Dev
Sum Minimum Maximum
y
30 8.38267 0.68124 251.48000
7.10000
9.52000
x1
30 3.73500 0.09016 112.05000
3.55000
3.90000
x2
30 3.94833
SAS Output
Page 1 of 9
regression
The REG Procedure
Model: MODEL1
Dependent Variable: hours
Number of Observations Read 17
Number of Observations Used 17
Analysis of Variance
Source
DF
Model
Sum of
Squares
Mean
Square F Value
3 489157494 163052498
Error
1
Baruch College, Statistics & CIS Dept.
Spring 2016
Dr. Yitzchak Sabban
Statistics 3155 EMWA - Course Outline
January 29 to May 28, 2016
Textbooks: FORECASTING, TIME SERIES, AND REGRESSION: AN APPLIED APPROACH
Fourth Edition by Bruce L. Bowerman, Richard T
Christina Hanchi
Professor Sabban
STA 3155
Chapter 7 Homework
Exercise 7.3
Both the multiplicative and additive yhats are shown in the table. The point forecasts should be able to
predict previous sales. The point forecasts using the multiplicative method
Department of Statistics and Computer Information Systems
Baruch College, City University of New York
Final Examination- Course: STA 3155, Weightage 40%
Instructor: Balwant S. Gill
Last Name: _ First Name: _ ID: _
Notes: 1. Support your answers with calcu
Department of Statistics and Computer Information Systems
Baruch College, City University of New York
Second Test: Course: STA 3155 Time: Take Home Test
Instructor: Balwant S. Gill
Last Name_ First Name _ID_
Notes: Support answers to Q2-Q5 by calculations
8. Exponential Smoothing
This forecasting method is most effective when the components (trend
and seasonal factors) of the time series may be changing over time
8.1 Simple Exponential Smoothing:
No trend model yt = 0 + t may be used to describe time serie
Department of Statistics and Computer Information Systems
Baruch College, City University of New York
Second Test: Course: STA 3155 Time: Take Home Test
Instructor: Balwant S. Gill
Last Name_ First Name _ID_
Notes: Support answers to Q2-Q5 by calculations
Department of Statistics and Computer Information Systems
Baruch College, City University of New York
Final Examination- Course: STA 3155, Weightage 40%
Instructor: Balwant S. Gill
Last Name: _ First Name: _ ID: _
Notes: 1. Support your answers with calcu
Chapter 5
Model Building and Residual Analysis
Model building refers to an approach to develop concepts and procedures for making decision to include
variables in the model, exclude influential data that deteriorate the prediction and inference and invest
Chapter 6
Time Series Regression
Time series regression models relate dependent variable to functions of time.
These models are most profitably used when parameters describing the time
series to be forecast remain constant over time. e.g if a time series
7. Decomposition Methods
The major goals of every time series forecasting are to identify and isolate the influencing factors for
predictive purposes. Models that help to decompose the time series into several factors: (trend, seasonal,
cyclical and irreg
9. Nonseasonal Box-Jenkins Models and Their Tentative Identification
Four Steps of Box-Jenkins Methodology:
1.Tentative Identification: Searching appropriate model using historical data.
2.Estimation: Estimation of parameters of the identified model
3. Di
Chapter 4
Multiple Linear Regression
4.1. The Linear Regression Model:
General Linear Regression Model:
y = 0 + 1x1 + 2x2 + 3x3 + + kxk +
4.2. The Least Squares Estimates, and Point Estimation and Prediction
A point prediction of an observed value of dep