Professional Documents
Culture Documents
Introduction
This paper compares the performance of trip generation
models. Trip generation, which is the first phase of
conventional four-step travel forecasting framework, estimates the number of trips to and from each traffic analysis
zone (TAZ) for various purposes. This trip-based traditional approach is still the standard practice for most
strategic transport planning, even though advanced
approaches like tour- and activity-based models explore
more realistic representations of behavior in travel demand
studies (McNally and Rindt, 2007; Donnelly et al., 2010).
78
2014
VOL
NO
Chang et al.
2014
VOL
NO
79
Chang et al.
Regression-type models
qn ~qn
if qn 0
if
(2)
qn w0
Regression model
Regression analysis in trip generation functionalizes the
relationship between trip generation rates, or the dependent variable, and a set of independent variables
qn ~bT xn zen ,
n~1,2, ,N
(1)
Poisson model
Trip rates, or the dependent variable, clearly show a
discrete nature. In addition, a very large proportion of trip
rates by households are zeros or small values. Regression
analyses based on least squares hold limitations for
dealing with these characteristics. A count data model
such as a Poisson regression can be an alternative
framework
Tobit model
Only the part of the distribution above q 50 is relevant to
the modeling for trip generation. Conventional regression
methods fail to account for the qualitative difference
between limit (zero) and non-limit (continuous) observations. The tobit model is a useful tool to deal with these
censored data. When data are censored, the distribution
that applies to the sample data is a mixture of discrete and
continuous distributions (Cotrus et al., 2005). To analyze
this distribution, a new random variable q which is
transformed from the original q is defined
80
e{ln lqnn
, n~1,2, ,N
P qn ~qn ~
qn !
(4)
2014
VOL
NO
Chang et al.
(5)
if hk1 vqn hk
..
.
Cross-classification-type models
Category analysis
The category analysis, or the cross-classification method,
would be the most extensively used approach for trip
generation. In this framework, household types are
classified according to a set of categories that are highly
correlated with trip-making. The dependent variable is
assumed to be continuous and two or three explanatory
variables, each broken into three or four discrete levels,
are usually applied. The independent variables can include
household size, car ownership, household income, and
some measures of land-use.
Mathematically, each household category constitutes a
cell in the cross tabulation. The average trip rate in a cell is
then calculated by simple algebra
Qh
q- h ~
H
(8)
Multiple classification
The conventional category analysis suffers from two
important methodological drawbacks: a non-statistical basis
and cell-by-cell calculation. The multiple classification
method can overcome these disadvantages. A statistically
justifiable approach based on a series of ANOVAs is applied
to select independent variables and their strata. First, oneway ANOVAs between trip rates and each candidate
variable are used to find the best grouping for each variable.
In this process, measures such as an F statistic and R2 can be
used. Once statistically significant independent variables
have been identified, multi-way ANOVAs between trip
rates and two or three candidate variables are applied. It is
an extensive trial-and-error procedure to examine the best
classification scheme. The eta-square g2, which is the ratio
of the sum of squares of the candidate variable to the
corrected total in the ANOVA output, can be used to
determine more contributable variables (Stopher and
McDonald, 1983). A grand-group mean approach is used
to estimate the cell value. The grand mean is calculated over
the entire sample, while the group mean is computed from
the row and column sums of the category table. A cell value
is found by adding the deviations of the cell to the grand
mean; this process is different from the cell-by-cell
calculation in the conventional category analysis (Stopher
and McDonald, 1983).
There are two drawbacks to multiple classification analysis. The ANOVA-based statistically justifiable approach
requires extensive trial-and-error procedures with no claim
of optimality. Its empirical acceptance is subject to low
efficiency in selecting good classification structures. The
grand-group mean calculation also loses the class-to-class
relationship between independent variables since one
variable is computed over all classes of the other variable
(Stopher and McDonald, 1983).
2014
VOL
NO
81
Chang et al.
models since the scales are different. The same holds true
for the standard goodness-of-fit measures such as the Ftest, R2, and t-test.
The performance comparison between models is conducted with validation. Validation tests a models ability to
predict future behavior. It requires comparing the model
output with information other than that used in estimating
or calibrating the model. Namely, the model output is
compared with observed travel data. Two ways of checking
model performance are considered in this paper. They are
the historical method and cross-validation.
Cross-validation is a statistical method for validating a
predictive model. Subsets of the data are held out for use
as validating sets; a model is fit to the remaining data and
used to predict for the validation set. Averaging the
quality of the predictions across the validation sets yields
an overall measure of prediction accuracy.
There are three distinct forms in the cross-validation.
Table 1 summarizes the advantages and limitations of the
cross-validation techniques. The simplest type would be
the hold-out technique, which is adopted in this study. It
divides observations into two subsets: one is for estimation and calibration, and the other is for validation.
Another form of cross-validation leaves out a single
observation at a time; this is similar to the jackknife
technique in the resampling. Lastly, the K-fold crossvalidation technique splits the data into K subsets; each is
held out in turn as the validation set (Picard and Cook,
1984; Kohavi, 1995).
The hold-out method is easy to apply, but there are few
reliable rules for classifying samples into the estimation
and validation sets. If data for estimation and validation
are sufficient enough as in this study, this issue does not
empirically matter much. The leave-one-out technique
may incur minimal errors because it simulates total
sample-size times, leaving just one observation out in
every trial. However, this method is inapplicable to large
sized datasets. The K-fold method requires acceptable
computation time and may generate reasonable errors in
the validation. However, this scheme also suffers from the
arbitrariness in determining the number K.
The historical technique is either forecasting, in which a
prior-year model is used to forecast current travel that is
then compared with actual current travel, or back-casting,
in which a current year model is used to estimate travel for a
prior year that is then compared with actual travel in the
prior year. The literature establishes little consensus on the
better of the two: the choice is fundamentally datadependent (Committee for Determination of the State of
the Practice in Metropolitan Area Travel Forecasting,
2007). Using both trials, of course, would be good practice.
In this paper, back-casting is used due to data availability.
Measures
It would be desirable to have some quantitative measures
for validation. In this paper, measures of correlation,
variance, and coincidence are considered (Cambridge
Systematics, Inc., 2010).
Correlation refers to a statistical relationship between
the observed and estimated trip rates. The closeness can be
expressed numerically by the correlation coefficient r
sq q^
(9)
r~ n n
sqn sq^n
where ^
qn refers to the estimated value of qn , sqn q^n is the
qn , and sqn and sq^n are the
covariance between qn and ^
qn , respectively.
standard deviations of qn and ^
Root-mean-square error (RMSE) is a measure of the
differences between values predicted by a model and those
actually observed
v
u N
u1 X
2
^
(10)
qn {qn
RMSE~t
N n~1
Though RMSE is a good index of model accuracy, it is
vulnerable to the scale problem. Per cent RMSE is an
alternative because it is the normalized version of RMSE,
removing scaling effects
v
u N
u1 X ^
qn {qn 2
(11)
% RMSE~t
N n~1
qn
Low RMSEs and high values of the correlation
coefficient are only one desirable measure of fit. Another
important criterion is how the model simulates the pattern
of the data. Namely, the distribution of estimated trip
frequency should be analogous to the shape in the data.
Thus, the ability to duplicate turning points or rapid
changes in the data is an important criterion for model
evaluation. The coincidence plot and its corresponding
ratio are a useful tool to address this issue. The plot
compares estimated and observed trip frequency distributions. The ratio measures the per cent of area that
coincides. This rate lies between zero and one, where zero
indicates two disjoint distributions and one means an
identical pattern (Cambridge Systematics, Inc., 2010)
(
!)
^k
P
H
Hk
min P
,P
^
k
k Hk
k Hk
(
!)
(12)
^k
P
H
Hk
max P
,P
^
k
k Hk
k Hk
Hold-out method
K-fold method
Leave-one-out method
82
Advantages
Limitations
Easy to apply
Known to involve reasonable levels of errors
Acceptable computational needs
Minimal errors incurred in principle
2014
VOL
NO
Chang et al.
Empirical analysis
Data
The study area is the Seoul metropolitan region in Korea.
The geographical sector includes Gyeonggi province;
Incheon, which is the fourth largest city in Korea; and
Seoul, the largest city and capital of the country. Subregions establish the functional relationships in terms of
lifestyles, economic activities, urbanization, and land-uses.
There are two broad sets of data used in this study. The
data for household characteristics and trip generation rates
are based on the household travel diary survey of the Seoul
metropolitan area. The Korea Regional Development
Total Information System (REDIS) database supplies the
regional characteristics that affect home-based work trips.
The Korea Regional Development Total Information
System is the official statistical database of the Korea
Presidential Committee on Regional Development. It
collects more than 300 regional statistics and has provided
the data to the public since 2008. This study has used four
types of data: demographics, regional economy, transport
system, and land-use.
The travel diary survey was started in 2002 by the code
prescribed in the Korea Intermodal Surface Transportation
Efficiency Act. This study extracted home-based work trips
from data surveyed in 2002 and 2006. The data of 211 564
households in 2006 and those of 159 068 households in
2002 were collected, as shown in Table 2. Two-thirds of the
data from 2006 were used for the estimation while the
remaining data from 2006 and all the samples in 2002 were
applied for the validation. Since the sample size of the 5 or
more trip rate is too minimal compared to those of the
others, the category was excluded in this study. Thus the
sample size in Table 5 is 148 018 not 148 091.
Estimation results
Table 3 shows the independent variables considered in the
study. Three kinds of household characteristics and four
sets of regional characteristics were considered.
The specifications for regression, tobit, Poisson, and
ordered logit models are shown in Table 4. These are the
Data in 2002
Trip rate
Total
Estimation set
Validation set
Validation set
0
1
2
3
4
5 or more
Total
Average
Standard deviation
43 521 (20.57)
115 358 (54.53)
43 714 (20.66)
7630 (3.61)
1234 (0.58)
107 (0.05)
211 564 (100.00)
1.1130
0.7702
30 455 (20.57)
80 738 (54.52)
30 616 (20.67)
5337 (3.60)
872 (0.59)
73 (0.05)
148 091 (100.00)
1.0928
0.7804
13 066 (20.59)
34 620 (54.54)
13 098 (20.64)
2293 (3.61)
362 (0.57)
34 (0.05)
63 473 (100.00)
1.0920
0.7799
29 237 (18.38)
91 139 (57.30)
31 504 (19.81)
5961 (3.75)
1126 (0.71)
101 (0.06)
159 068 (100.00)
1.0926
0.7802
The number in parentheses refers to the percentage; return-home trips are excluded since origin-destination-based data are only available
in 2002 while production-attraction-based data are also built in 2006.
2014
VOL
NO
83
Chang et al.
Household
characteristics
Variable type
Variables considered
Household composition
Number of telecommuters
Number of employees
Number of household members
Number of preschool children
Household income
House ownership
Car ownership
Number of cars owned
Subway availability
Number of household members with driving licenses
Total population
Population density
Population moved in
Population moved out
Annual expenditures
Annual expenditures per capita
Annual expenditures per capita per area
Annual tax revenues
Annual local tax revenues
Annual local tax per capita
Fiscal self-reliance ratio
Number of car registered
Number of car registered per capita
Number of subway stations
Number of subway lines
Number of bus routes
Road density
Gross area
Residential area
Commercial area
Industrial area
Green area
Number of employees
Number of employees per area
Number of companies
Number of companies per area
Income level
Transportation
Regional
characteristics
Demographics
Regional economy
Transport system
Land-use
Table 4 The specication for regression, tobit, Poisson, and ordered logit models
Variable
Dependent variable
Trip rates
Independent variables
Number of employees
Household income
Car ownership
Population density
Annual expenditures per capita per area
Number of subway lines
Number of companies per area
84
Specification
2014
VOL
NO
Chang et al.
Table 5 Estimation results for regression, tobit, Poisson, and ordered logit models
Constant
Number of employees
Household income
Car ownership
Population density
Annual expenditures per capita per area
Number of subway lines
Number of companies per area
Regression
Tobit
Poisson
0.0098
0.5394***
0.0947***
0.0072*
0.0022***
0.3014***
0.0091***
0.0551***
20.3352 (20.3073)***
0.6166 (0.5652)***
0.1303 (0.1194)***
0.0323 (0.0296)***
0.0031 (0.0028)***
0.2972 (0.2724)***
0.0101 (0.0093)***
0.074 (0.0678)***
0.7248 (0.6644)***
0.9167
20.9518
0.4208
0.1004
0.0276
0.0030
0.2878
0.0068
0.0432
148 018
MLE
148 018
MLE
3.5167***
6.4983***
9.0627***
148 018
MLE
75 740.6***
29 981.8***
77 142.6***
Ordered logit
(21.0383)***
(0.4591)***
(0.1095)***
(0.0301)***
(0.0033)***
(0.3139)**
(0.0075)
(0.0471)***
21.8605***
1.9776***
0.3307***
0.0469***
0.0063***
0.9318***
0.0314***
0.1880***
1.0909
h1
h2
h3
Sample size
Estimator
F-statistic
^
{2 L0{L b
148 018
OLS
68 367.6***
R2
0.4005
0.0828
0.2302
^
* 10% significance level; ** 5% significance level; *** 1% significance level; {2 L0{L b is the likelihood ratio statistic where L(0) is
^
the value of the log likelihood function when all parameters are zero and L b is that at its maximum.
2
r
0.1765
Medium
High
01000 K won
10002000 K won
20003000 K won
3000 K won or more
01000 K won
10002000 K won
20003000 K won
3000 K won or more
01000 K won
10002000 K won
20003000 K won
3000 K won or more
3 or more
0.26
0.69
0.87
0.90
0.08
0.26
0.32
0.41
0.11
0.33
0.39
0.47
0.47
0.76
0.84
0.86
0.73
0.85
0.89
0.90
0.74
0.85
0.91
0.92
0.76
1.27
1.48
1.58
1.35
1.54
1.61
1.66
1.36
1.53
1.60
1.69
1.37
1.52
1.98
2.27
1.83
2.11
2.22
2.39
1.93
2.18
2.33
2.51
q50
q51
q52
q53
q54
Number of employees
Household income
Car ownership
Population density
Annual expenditures per capita per area
Number of subway lines
Number of companies per area
20.20249
20.03386
20.00480
20.00065
20.09540
20.00322
20.01925
20.09550
20.01597
20.00227
20.00031
20.04500
20.00152
20.00908
0.27576
0.04611
0.00654
0.00088
0.12993
0.00438
0.02621
0.02048
0.00342
0.00049
0.00007
0.00965
0.00033
0.00195
0.00175
0.00029
0.00004
0.00001
0.00082
0.00003
0.00017
2014
VOL
NO
85
Chang et al.
Table 8 The result of one-way ANOVAs for the multiple classication analysis
Candidate variable
Stratification
F-statistic
Number of employees
Household income
Car ownership
Population density
Annual expenditures per capita per area
Number of subway lines
Number of companies per area
0, 1, 2, and 3 or more
01000 K won, 10003000 K won, and 3000 K won or more
0 and 1 or more
Low (05000), medium (500018 000), and high (18 000 or more)
0100, 100200, 200300, and 300 or more
0, 1, 2, 3, and 4
0500, 5001000, and 1000 or more
31541.922***
4647.556***
1306.429***
722.164***
58249.731***
33.762***
285.863***
Population density, annual expenditures per capita per area, and number of companies per area are represented by persons per square
kilometer, 1000 won per capita, and companies per square kilometer, respectively; and *** 1% significance level.
Medium
High
F-statistic
2
R
86
01000 K won
10003000 K won
3000 K won or more
01000 K won
10003000 K won
3000 K won or more
01000 K won
10003000 K won
3000 K won or more
3237.538***
0.434
3 or more
0.07
0.52
0.66
0.18
0.36
0.43
0.31
0.50
0.59
0.10
0.54
0.68
0.73
0.92
0.99
0.82
1.00
1.10
0.69
1.13
1.27
1.45
1.64
1.71
1.53
1.71
1.81
1.18
1.63
1.77
2.09
2.27
2.34
2.25
2.44
2.53
Performance comparison
The performance comparison between the models is
conducted based on validation. As stated in the methodological section, both cross-validation and back-casting
were adopted in this paper. The measures of correlation,
variance, and coincidence were utilized for the quantitative assessment of this process.
Since validation is basically represented by the comparison between the observed and modeled values, the first
task is to determine estimated trip rates for each model.
This is straightforward in the case of discrete choice
models. The continuous and count data models, however,
require an auxiliary step to convert model outputs to
integer approximations. The category-type models also
need this additional phase. The procedure adopted in this
paper is to round off the values to the nearest integer. This
process, however, inevitably involves biases.
Table 10 summarizes the results of validation. In the
cross-validation, the models show similar RMSEs while the
category-type approaches achieve slightly better performance. This trend is also observed from the % RMSE
measure. This result is not surprising because the two
indices represent fundamentally the same performance, or
variance, even though % RMSE eliminates the scale
problem of RMSE. The differences in model performance
are noticeable when viewing the correlation indicator.
Again, the category-type methods accomplish a better
goodness-of-fit while the Poisson model records the worst.
A comparable finding is seen in the coincidence index.
Back-casting results in consistent but somewhat different
outcomes from that of cross-validation. For the criteria of
variance and correlation, the traditional regression analysis
shows better performance. Again, the count data model
performs the worst. The most obvious difference in the
model performance can be found from the measure of
coincidence. This index has been conceived to identify a
models ability to simulate patterns in the data. The
category-type models achieve remarkable performance in
this measure.
Thus, it can be summarized that the category-type
models show superior performance. The regression model
also generated an acceptable level of performance. Even
2014
VOL
NO
Chang et al.
Cross-validation
RMSE
% RMSE
Correlation coefficient
Coincidence ratio
Back-casting
RMSE
% RMSE
Correlation coefficient
Coincidence ratio
Regression
Tobit
Poisson
Ordered logit
Category
MCA
0.6787
62.16%
0.5405
0.7531
0.6808
62.34%
0.5402
0.7407
0.6865
62.86%
0.4899
0.5640
0.6812
62.38%
0.5471
0.7713
0.6423
58.82%
0.6107
0.7500
0.6466
59.22%
0.5975
0.7430
0.6126
55.04%
0.6241
0.6885
0.6391
57.42%
0.6037
0.6805
0.6548
58.84%
0.5389
0.5387
0.6499
58.39%
0.5764
0.6971
0.6456
58.01%
0.5761
0.8188
0.6478
58.21%
0.5704
0.8134
though this result does not directly mean that the betterperformance model should become the standard framework for trip generation forecasting, the conventional
models can be said to address a satisfactory trip-making
behavior.
It should also be noted that the more sophisticated
models have not shown better performance in terms of
validation, though their implementation is more onerous.
This does not mean that the models do not provide more
advanced frameworks. Indeed, the models have more
appealing theoretical bases compared to those of the
conventional approaches. It is difficult, however, to find
ways to improve the mechanism of trip-making behavior
of individuals against the traditional methods. It is one of
the causes for the disappointing performance of the
advanced models.
Conclusion
Trip generation forecasts the number of trips that begin
from or end in each travel analysis zone. This modeling is
the first phase of the four-step travel forecasting procedure. Traditionally, regression and category analyses have
been applied to this step and have generated an acceptable
level of performance from the perspective of planning.
However, structural limitations for each type have also
been criticized. The negative trip rate likelihood, the
continuous nature in trip rates, and the lack of incorporation of traveler behavior characteristics are the typical
problems for the regression model. The category analysis
has also suffered from the drawbacks of non-statistical
justification and cell-by-cell calculation of trip rates.
Several alternative approaches have been put forward,
but a systematic investigation into the performance
between the models had not sufficiently been performed.
This paper provides such an analysis, using the Seoul
metropolitan region as the study area. The household
travel diary data and REDIS were used for modeling. Six
kinds of models were estimated and validated. The results
show that the category-type models are superior in overall
performance.
The results of this kind of comparative study may be
specific to the datasets used. Some approaches may lead to
better replication of observed patterns; however, they may
not lead to better forecasts. Namely, the findings by this
study should not be understood to give a green light to the
traditional methodologies for being the standard framework
Acknowledgements
This work was supported by the BK 21 plus program of
the National Research Foundation of Korea.
References
Agyemang-Duah, K. and Hall, F. L. 1997. Spatial transferability of an
ordered response model of trip generation, Transp. Res. A Policy
Pract., 31, 389402.
Badoe, D. A. 2007. Forecasting travel demand with alternatively
structured models of trip frequency, Transp. Plann. Technol., 30,
455475.
Bastida, C. and Holguin-Veras, J. 2009. Freight generation models,
Transp. Res. Rec. J. Transp. Res. Board., 2097, 5161.
Barmby, T. and Doornik, J. 1989. Modelling trip frequency as a Poisson
variable, J. Transp. Econ. Policy, 309315.
Cambridge Systematics, Inc. 2010. Travel model validation and reasonableness checking manual, 2nd edn, Travel Model Improvement
Program, Federal Highway Administration, Washington DC, US.
Committee for Determination of the State of the Practice in Metropolitan
Area Travel Forecasting. 2007. Metropolitan travel forecasting
current practice and future direction, Special Report 288,
Transportation Research Board, Washington, DC.
Cotrus, A., Prashker, J. and Shiftan, Y. 2005. Spatial and temporal
transferability of trip generation demand models in Israel, J. Transp.
Stat., 8, 3756.
Daly, A. 1997. Improved methods for trip generation, Transportation
Planning Methods Volume 11, Proceedings of Seminar F held at PTRC
European Transport Forum, Brunel University, England, 207222.
Donnelly, R., Erhardt, G., Moeckel, R. and Davidson, W., 2010.
Advanced practices in travel forecasting, National Cooperative
Highway Research Program Synthesis 406, Transportation
Research Board, Washington DC, US.
Greene, W. H. 2000. Econometric analysis, 4th edn, London, Prentice
Hall International, Inc.
Jang, T. Y. 2005. Count data models for trip generation, J. Transp. Eng.,
131, 444450.
Kohavi, R. 1995. A study of cross-validation and bootstrap for accuracy
estimation and model selection, the 14th International Joint
Conference on Artificial Intelligence, Montreal, Quebec, Canada,
11371145.
Lim, K. and Srinivasan, S. 2011. Comparative analysis of alternate
econometric structures for trip generation models, Transp. Res.
Rec., 2254, 6878.
Ma, J. and Goulias, K. G. 1999. Application of Poisson regression models
to activity frequency analysis and prediction, Transp. Res. Rec. J.
Transp. Res. Board, 1676, 8694.
McNally, M. G. and Rindt, C. R. 2007. The activity-based approach, in
Handbook of transport modelling, (eds. D. A. Hensher and K. J.
Button), Amsterdam, Netherlands, Elsevier Science Ltd, pp. 5574.
2014
VOL
NO
87
Chang et al.
88
2014
VOL
NO