Upload
vandang
View
218
Download
2
Embed Size (px)
Citation preview
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 655
Application of Forecasting Technique for Students Enrollment
�orhaidah Abu Haris1,Munaisyah Abdullah
2, Abu Talib Othman
3and Fauziah Abdul Rahman
4
1Universiti Kuala Lumpur (U*IKL), Malaysia,[email protected] 2Universiti Kuala Lumpur (U*IKL), Malaysia,[email protected]
3Universiti Kuala Lumpur (U*IKL), Malaysia,[email protected] 4Universiti Kuala Lumpur (U*IKL), Malaysia,[email protected]
ABSTRACT
Students enrollment forecasting provides
information for management in decision making and
planning for the Higher Education Institutions
(HEI). Due to that, many forecasting techniques
have been proposed to improve the accuracy in the
predictions. This study discussed the factors that
influence the students’enrollment and various
forecasting techniques that have been applied in
several studies.The understanding offorecasting
techniques is important to find which techniques
could give the best result and could contribute
informative knowledge, especially for management
to make decisions in HEI students' enrollment area.
Keywords:enrollment, forecasting, decision
making.
I I�TRODUCTIO�
Forecasting is an important tool for both strategic
and tactical decision making that is considered as an
essential part for efficient and effective
management. The reasons for management to
considerforecasting are to study the current,
historical and the continuance of data relationships
for certain situations, where that forecasting
techniques that will providea high level of accuracy
with less cost, the quantity of the historical data in
which the relevancy of the information can be
measured for reliable decision making and lastly,
the time frame to formulate the forecast.(Dianne &
Amrik S., 1994)
Forecasting will help management to make further
assumptions based on data collected. Enrollment
forecasting is related to analyzing current trends,
understanding the significant impact on the
enrollment and revenue outcomes. This information
could be used to influence the future strategy and
resource decisions. (Ward, 2007)
Student enrollment forecastingprovides important
information for management decision making and
planning in Higher Education Institution (HEI).
The significance of this study is tounderstand
various factors that influencestudents to enroll in
HEI. Two forecasting techniques are identified in
this study – Statistical and Data Mining shows that
each technique has their own strengths and
weaknesses to produce forecast accuracy.Some of
the influential factors have been used by the
researchers to forecast the students’ enrollment
decisions in HEI. The selection of HEI for students’
enrollment could be varied between different HEI
and different countries. Therefore, factors contribute
in influencing the students to enroll also could be
deferred.
Therefore, some of the knowledge contribute from
the chosen forecast techniques is dependent on the
selection factors of the particular HEI.
II FACTORS I�FLUE�CE STUDE�TS
E�ROLLME�T I� HEI
Studies on students’ decisions have attracted many
researchers. The influencingfactorsthat lead toa
conclusion on the student enrollment decisions can
be divided into three main factors that can be further
categorized into eleven sub-factors. The factors are:
(Farhan et al, 2012)
a. Internal Factors – Aspiration, Aptitude and
Career
b. External Factors – Courses, Cost Location,
Reputation, Promotion and Facilities.
c. Social Factors – Parents, Friends and Teachers.
The factors influencing the students’ choice to study
in HEI are important because HEI will be able to
understand the reasons why a student chooses a
particular institution over others. The information
obtained can be used by universities to assist in the
development of a marketing plan.(Mahsood &
Chenicheri, 2010) Therefore the management
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 656
couldlayout strategic planning to promote their HEI
to potential students’.
Other factors that could influence the students’
enrollment decisions in a particular HEI are:
a. Financial Aid - financial aid merit awards
provided by the HEI.(Klaauw, 2002)(James,
2009)(Eric et al, 2009), the availability of
financial aid can affect enrollment and
completion(Deland & Cochrane, 2008). The
importance of financial aid is an effective tool to
compete with other colleges to obtain
students.(Klaauw, 2002) The types of financial
aid – merit based scholarships, need based
scholarship and educational loan(Janet, 2007).
The choice of public or private depends on the
financial aid offered for students from low
income families.(Hashem & John A., 2010)
b. Information Satisfaction - the information
regarding product or service provided by the HEI
(Farhan et al, 2012).Students’ enquiry on the
institutions can also predict the enrollment
decision(Monks, 2009).
c. Employment Opportunities – employment
opportunities upon completion of studies.
(Klaauw, 2002).
d. University or College choice – depends on the
facilities, procedures and policies, entry
requirement and extra-curricular
activities.(Samsinar et al , 2003)(Wagner &
Pooyan, 2009) and quality of services
provided.(Wagner & Pooyan, 2009)
e. Tuition fee - Cost of education and degree
(content and structure) (Wagner & Pooyan,
2009)
f. Institution Prominence – The reputation of the
HEI and desired programmes(Fernandez, 2010),
quality of education and also pecuniary factors.
(Fernandez, 2010)
g. Student Characteristics, External Influences and
College Attributes(Ming, 2010)
h. Location - Geographic and demographic factors
also influence the students' enrollment decision
in HEI. (Monks, 2009)
Besides all the factors mentioned above, the main
reasons students pursue higher education were to
improve job prospects, gain knowledge and
experience. (Fernandez, 2010).Other factors that
could influence student’s decisions are facilities and
infrastructure (Ward, 2007).The studentinquiries in
choosing the HEI of their choice can also be used as
a predictor.(Cullen F. & Kenton, 2006). (Ming,
2011) in his studies emphasize that students from
different countries have different criteria in selecting
HEI. He also introduces a conceptual framework
which shows the factors that influencing student
decision to enroll in HEI in Malaysia.(Ming, 2010)
Figure 1. Conceptual Framework (Ming,2010)
In figure 1, the conceptual framework consistsof independent variables and dependent variables.The independent variables are divided into two: Fixed College Characteristics and College Effort to Communicate with students.
III FORECASTI�G TECH�IQUES O�
STUDE�T E�ROLLME�T
Forecasting techniques can be classified as either
extrapolation– extend present trends or normative –
look backward for a desired future.Technology for
forecasting techniques can be categorized
into:(Roper et al, 2011)
a. Direct – forecast parameters that measure
functional capacity or relevant characteristics.
b. Correlative – correlate parameters that measure
technology with parameters of other
technologies or other background parameters.
c. Structural – interaction between technology and
context that must be explicit.
The enrollment forecasting methods are critically
important for the HEI to determine short and long
term forecasts for capital improvement and facilities
planning. Accurate forecasts are important in times
of change. The increasing and decreasing number of
students will not only affect planning decisions, but
could givemanagement perceived quality of
instructions.(Sweeney & Middleton, 2005)
Various techniques used by HEI to forecast
enrollment. The forecasting techniques are divided
into a few categories: (Ahrens, 1979) (i) quantitative
techniques based on historical data (ii) contribution
models that incorporate the historical data and rely
on a relationship between enrollments with other
parameters, or using techniques that integrate
subjective judgment instead of using quantitative
measures; (iii) modification or suggestionsfor
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 657
adjustments or comparisons to previously developed
forecasting techniques (iv) qualitative techniques
using surveys to determine potential students’
enrollment(Nurlida, Faridah , & Nooraini, 2013).
However the selection of appropriate forecast
techniques is dependent on the availability of the
HEI data. The techniques involved in each
forecasting will give input on the long and short
term range planning of the HEI.
The field areas that are identified for HEI forecast:
(Ward, 2007)
a. Size Scope – number of students projection
for particular year or semester, target goal
achievement for HEI, influence on student
credit and tuition revenue.
b. Profile on the Students Enrollment – the
number of students that pursuing degree and
completion.
c. New Student academicprofile ability -
entrance exam scores, grade point average
and etc.
d. Enrollment Mix – gender, ethnicity,
international status, registration status (full
time vs part time), lodging status
(residential vs commuter)
e. Enrollment Outcome – persistence rate for
first time student enrollment enhancement,
student retention sustained or improved.
f. Financial Aid and Revenue – can support
the size, enrollment mix and enrollment
outcomes.
The need for forecasting as explained by (Huang &
Yu, 2013)for situations such as:
a. Difficult for decision makers to evaluate the
complexity of organizational and environmental
factors without systematic assistance.
b. The organization's relationship could be not
reliable, therefore, need forecasts to identify new
relationship.
c. Organizations could be moving toward systematic
decision making therefore need formal
forecasting to support and assess their activities.
d. Practitioners could have developed and directly
applied formal forecasting methods.
Therefore, forecasting will enable decision makers
to focus on the variety and complexity of the
problem. The selection of the correct technology
forecasting techniques for a particular problem will
help decision makers to understand better
possibilities. The following methods have been
identified to forecast on the student enrollment in
HEI and theresults weredependent on HEI
forecasting criteria.
A. Statistical
A few statistical methodshave been used to forecast the student enrollment in HEI with varyingresults.
Structural Equation Modelling (SEM) is used to identify the factors that are responsible for student choice in private HEI - experienced or qualified or knowledgeable lecturers, suitable syllabus. The finding has shown that the academician’s expertise and promotion medium will attract students to choose HEI for their educational studies. (Osman M. Zain et al, 2013)
Forecasting using Regression Analysis was applied tothe number of students enrolled by semester and level, freshman and senior as well as student demand for theme courses seats in the general education program in Grand Valley State University. The results on the forecasted historical data has been used to change on the academic policy. (Shabbir Choudhuri et al, 2007)
Institutional Research Intelligence (IRI) uses SAS Logistic Regression to predict the students that were offered admissions to HEI but turned down the offer. (Djunaidi, 2012)
The Monte Carlo method is used to forecast the
enrollment for KIMEP University, Kazakhstan. In
the studies, the findings mention that the student
choice satisfaction depends on the information
satisfaction and the information satisfaction has no
direct impact on institution choice
satisfaction.(Quinn, 2013)
Multiple Regression Analysis was used in the study
hadfocused on the Internal, External and Social
factors that influence students' choice decisions to
enroll in HEI in Pakistan. The objective of the study
is to explore factors that influence students'
decisionsby applying multiple statistical techniques
to find out association between them. (Farhan et al,
2012)
The study conducted by (Shuqin, 2002) examined
three enrollment projection methods and tested on
six community colleges and compared using Mean
Absolute Percent Error (MAPE). These methods are
Linear Regression, Auto-Regression and Three
Component Model that was applied using different
variables and which results islesser percentage of
error.Linear Regression was applied to student
population, college budget and student fees. Auto-
Regression was applied to population of ages 16 to
55, college budget and student fees. The Three
Component Model was tested on four groups of new
students (first time and non-credit students)with the
age population that are between the ages of 18-19 ,
20-24, 25-34 and 35-55.The results of the studies
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 658
show that the three methods did quite well and the
error rate is low.
Mindanao State University (MSU-IIT) used
students’data to forecastthe number of courses to be
opened before the enrollment commences. This
study was conducted using statistical time series
models – Simple Moving Average, Single
Exponential Smoothing and Double Exponential
Smoothing. The results of the time series models
were compared with naïve model used by the
university and the projected data yields less error
using MAPE with 20.5% difference in favor of the
time series models.(Rabby Q. & Mary Jane, 2012)
All these methods were used to forecast the
enrollment of students in HEI with different criteria
and variables. These methods are satisfied for
certain conditions only.
B. Data Mining
Data Mining(DM) can also be used for forecasting.A
few of data miningtechniques has been identified to
forecast the studentenrollment in HEI.
The techniques of DM as described by (Fadzillah &
Mansour Ali, 2011) could be either predictive and
descriptive. The predictive techniques is to forecast
about values of data values using known results
found from different data sources while descriptive
techniques identify patterns or relationships in data.
The Predictive techniques consist of Classification,
Prediction, Regression and Time Series analysis. The
Descriptive techniques compriseof Clustering,
Summarizations, Association Rules and Sequence
Analysis.
In the study for student enrollment at Sebha
University in Libya, the cluster analysis is used to
group the analyzed data of the undergraduate student
enrollment into clusters based on similarities
consists of genders,Libyan students, courses and
etc.The result of is used as a predictor experiment to
the predictive Analysis encompass of Neural
Network, Logistics Regression and Decision Tree
which are applied on student enrollment. The
outcomes of the studies show that Neural Network
has given better prediction accuracy on the student
enrollment.(Fadzilah & Mansour, 2009)
In other research conducted by a private university in
Thailand, Bayesian Network of DM methods had
been applied to predict student course registration in
a private university in Thailand. The data were
obtained from student enrollments together with a
GPA and grades of the subjects. The results were
used to advise students in planning to register
courses in each semester. (Pathom et al, 2008)
The study later was further enhanced by comparing
Bayesian Network with 3 other types of classifying
algorithms- C4.5, Decision Forest and NBTree using
the same variables GPA and grades of the subjects.
NBTree was concluded to give highest prediction
power. (Pathom et al, 2008)
Other DM methods which are used to predict on the
enrollment of students in the California State
University areSupport Vector Machine and Rule-
based predictive techniques.The students are
grouped into three categories: new students include
freshman and transfer, continued and returned
students. The analyzed student’s data include
population, employment, tuition and fees. The result
was analyzed by Cubist a tool to better understand
the enrollment projection.(Svetlana et al, 2006)
The research conducted by University of Lima, Peru
is to find the impact on the academic performance
based on the students’ enrollment. Two synthetic
attributes were used – course difficulty and potential
(grades in related course). C.45, K-Nearest Neighbor
(K-NN), Naïve Bayes, Bagging and Boosting were
applied to the data. The result suggests that Boosting
is the best accuracy predictor methods which was
further tested on the Student Performance
Recommender System with a learning engine with
proven outcomes.(Vialardi, et al., 2011)
Another DM technique to forecast on the academic
performance using enrollment data is Decision Tree.
The studied result focuses on the students’ strength
at the time of enrollment, therefore after the
enrollment process; HEI will attempt to improve the
quality of the student. The input will help the
instructors to take necessary actions for the slow
learners to improve in the university
examinations.(M Narayana & M., 2012)
(Borah et al, 2011)hadproposed a new attribute
Selection Measure Function (heuristic) on existing
C4.5 Decision Tree algorithm. The C4.5 gives
promising solution on the split information of the
enrollment process at the engineering college in
India.Another approach in the studies is to find the
accuracy rate between C5.0 Decision Tree algorithm,
and back propagation algorithm of the Artificial
Neural Network(ANN). The resultsshowed that the
ANNhas given the highestforecasting accuracy.
The study that was conducted at the University of
California suggested that DM techniques are much
better than statistical methods to assist HEI in
achieving enrollment goals. SAS Enterprise Miner
software is appliedto test the DM techniques, Neural
Network, Decision Tree, Logistic Regression and
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 659
Ensemble Models. From this study Neural Network
method is better than other methods(Tongshan).
Studies on enrollment forecasting also has been
conducted using Fuzzy Time Series methods. These
methods are:High Order Fuzzy Time Series. (Chen
& Hsu, 2004), Fuzzy Time Series and Genetic
Algorithms (Chen & Chung, 2006), Fuzzy Time
Series and Clustering Techniques (Kurniawan &
Chen, 2009), Adaptive Time Variant Fuzzy Time
Series(Wong, Enjian, & Chu, 2010), High Order
Fuzzy Time Series(S.M.Chen, 2002), First-Order
Fuzzy Time Series (Melike & Konstantin, 2005),
Heuristic Time Variant Fuzzy Time Series (Melike
& Konstantin, 2005), Cardinality Fuzzy Time Series
(Chang et al, 2007)were applied to find highest
forecasting accuracy.Another method is Weight
Fuzzy Time Series with additional data from
University Teknologi Malaysia(UTM).(Z. Ismail &
R.Efendi, 2011)
These Fuzzy Time Series techniques have been
conducted using enrollment data of theUniversity,
Alabama, which was originally introducedby Song
and Chissom.(Q.Song & B.S. Chissom, 1993)
Various DM forecasting techniques explained above
show that DM could be applied to improve the
accuracy inforecasting. The researchers have
proposed various techniques with different factors
and variables to suit with their forecast criteria. The
accuracy of each of the techniques is dependent on
various factors chosen for particular studies.
The studies are further categorized into the table as
shown in Table 1- the list of the Statistical methods
and Table 2 - the list of DM methods that was used
to forecast on student enrollment in HEI.
Table 1: Statistics
Statistical Methods Factors
Structural Equation
Modelling(SEM)
Academicians
Lecturers,
promotions
Regression Analysis
New+ existing
students,
courses
Monte Carlo Information
satisfaction
Multiple Regression
Internal,
External &
Social factors
Linear Regression
Student
population,
budget, fees
Auto Regression Student age,
budget & fees
Three Component
Model Student age
Statistical Time
Series –
a. Simple Moving
Average
b. Single
Exponential
Smoothing
c. Double
Exponential
Smoothing
Number of
courses to be
open
SAS – Logistic
Regression
Reject
admission
Table 2: Data Mining
Data Mining
Methods Factors
Neural Network Undergraduate
students
cluster(genders,
courses & etc)
Logistic
Regression
Decision Tree
Bayesian
Network
GPA & grades,
subjects – course
register
C4.5
Decision Forest
NB Tree
Support Vector
Machine (SVM)
New (freshman &
transfer),
continues&
returned
Selection
Measure
Function
(Heuristics) –
C4.5, Decision
Tree
Students
Enrollment
process Artificial Neural
Network – Back
propagation
Algorithm
SAS Enterprise
Miner – Neural
Network,
Decision Tree,
Logistic
Students
Enrollment
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 660
Regression,
Ensemble Model
Decision Tree Students’
Strengths
Fuzzy Time
Series (FTS),
High Order FTS,
FTS & Genetic
Algorithm,
FTS &
Clustering
Algorithm,,
Adaptive Time
Variant FTS.
First Order FTS,
Heuristic Time
Variant FTS,
Cardinality FTS,
Weight FTS
Number of
student
enrollments per
year
C4.5,
K- Nearest
neighbor,
Naïve Bayes,
Bagging &
Boosting
Course difficulty
& Grades
IV THE IMPORTA�CE OF
FORECASTI�G O� THE STUDE�TS
E�ROLLME�T
The knowledge learned from the forecasting
techniques could be important to explore the
potential of the organization’s historical data and
could be used to deal with management decisions of
the present situation(Jing, 2002).Knowledge
Management (KM) forecasting technologies could
also give significant impact on supporting the
technology of forecasting. These technologies
are:(Marc et al, 2006)
a. Data Mining – identified certain pattern of
information such as classification and
association analysis.
b. Case-Based Reasoning – support decisions and
solve problems.
c. Information Retrieval – target of the raw
information and the resulting information is
retrieved through references or alike or need to
be completely revised.
d. Topic Maps – provide methods to navigate
associative across large amounts of available
information in a conscious manner, enabling
systematic identification of information and
creation of new knowledge by the user.
e. Ontologies – related to meta-context approach in
which the explicit contextual information and
specific references have added to the information
resources.
These KM forecasting can be supported by one or more technology, which is part of strategic innovation management.
The problem of forecasting the data is very time consuming and usually involves various steps to define, extracted, cleaned, harmonized and prepared data. The reasons several techniques applied to these data is to produce a high quality forecast with best result. Understanding the factors that could drive the enrollment demand will help the HEI to derive waysto leverage numerous resources into actionable strategies that can directly impact profitability.
The choice of a good forecast technology could
either be quantitative, qualitative or mixture of both.
The rationale for selecting many different
approaches as practical within resource limitations
is to support the decision makers to make good
decision based onthe forecastinformation from a
range of desirable futures or avoid the less desirable
ones. This forecast technique will give impact on the
forecast assessment and at the same time could
affect new developments on the technologies,
processes, policies, organizations and other factors
in which forecaster can identify areas that would
give significant impacts that will occur, likelihood
and their consequences in future. ( Roper et al,
2011)
V CO�CLUSIO�
In this paper, a few methods are presented and
described in the literaturestudieson various research
that are related to forecastingon the student
enrollment in HEI. The forecasting techniques are
critical components for analyzing the consequences
and necessary to be applied in future works. Finally,
the ability to understand some of the forecasting
techniquesis important in finding the best and
suitable approach that could contribute
knowledgeable information especially in HEI
students'enrollment area.
ACK�OWLEDGME�T
We would like to express our thanks especially to
Universiti Kuala Lumpur (UniKL) for the first
author to take a PhD research.
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 661
REFERE�CES
Ahrens, S. W. (1979). Student Enrollment Forecasting
Techniques for Higher Education. Institute of
Education Science, 33.
Borah, M. D., R. J., Daya Gupta, & Deka, G. C. (2011).
Application of Knowledge Based Decision Technique
to Predict Student Enrollment Decision. 2011
International Conference on Recent Trends in
Information System, 180-184.
C. G., & K. P. (2006). A Predictive Model of Inquiry to
Enrollment. Research in Higher Education, 935-956.
Chang, J. R., Lee, Y. T., Liao, S. Y., & Cheng, C. H.
(2007). Cardinality Based Fuzzy Time Series for
Forecasting Enrollments. Springer-Verlag, 753-744.
Chen, S. M., & Chung, N. Y. (2006). Forecasting
Enrollments of Students by Using Fuzzy Time Series
and Genetic Algorithms. Information Management
Sciences, 1-17.
Chen, S. M., & Hsu, C. C. (2004). A New Method to
Forecast Enrollments using Fuzzy Time Series.
International Journal of Applied Science and
Engineering, 234-244.
D. C., & Cochrane, C. F. (2008). Paving the Way: How
Financial Aid Awareness affects College Access &
Success. Institute for College Access & Success.
D. W., & A. S. (1994). Forecasting: The Key to
Managerial Decision Making. Management Decision,
41-49.
E. B., Long, B. T., P. O., & L. S. (2009). The Role of
Simplification and Information in College Decisions:
Result from the H&R Block FAFSA Experiment.
*ational Bureau of Economics Research.
F. M., S. M., & N. B. (2012). Factors Influencing
Student's Enrollment Decisions in Selection of Higher
Education Institutions (HEI's). Interdisciplinary
Journal of Contemporary Research in Business, 558-
568.
F. S., & M. A. (2009). Uncovering Hidden Information
Within University's Student Enrollment Data Using
Data Mining. 2009 Third Asia International
Conference on Modelling & Simulation, 413-418.
F. S., & M. A. (2011). Mining Enrollment Data Using
Descriptive and Predictive Approaches,. In P. K.
(Ed.), Knowledge-Oriented Applications in Data
Mining (pp. 1-21). In-Tech.
Fernandez, J. L. (2010). An Exploratory Study of Factors
Influencing the Decision of Students to Study at
University Sains Malaysia. Kajian Malaysia Journal
of Malaysian Studies,28, no.2, 107-136.
H. D., & J. K. (2010). Enrollment at Highly Private
Colleges: Who is Left Behind? Contemporary
Economic Policy, 28(1), 94-109.
Huang, K. H., & Yu, T. K. (2013). Forecasting Regime
Switches to Assist Decision Making. Management
Decision, Emerald, 1-9.
J. C., S. M., & D. S. (1971, July). How to Choose the
Right Forecasting Technique. Retrieved March 2014,
from Harvard Business Review:
http://hbr.org/1971/07/how-to-choose-the-right-
forecasting-technique/ar/4
J. L. (2002). Data Mining and Its Application in Higher
Education. In *ew Directions for Institutional
Research. Wiley Periodicals.
J. R., & R. R. (2010). Applications of Data Mining
Techniques in Higher Education in India.
International Conference on Innovation in Redefining
Business Horizons (pp. 1-10). Journal of Knowledge
Management Practice, Vol 11.
J. W. (2007). Forecasting Enrollment to Achieve
Institutional Goals. College and University Journals,
41-46.
K. T., & Chen, S. M. (2009). A New Method to Forecast
Enrollments using Fuzzy Time Series and Clustering
Techniques. Proceedings of the Eighth International
Conference on Machine Learning and Cybernetics
(pp. 3026-3029). Baoding: IEEE.
K. W., & Fard, P. Y. (2009). Factors Influencing
Malaysian Students' Intention to Study at a Higher
Educational Institution. E-Leader Kuala Lumpur, 1-
12.
Klaauw, W. V. (2002). Estimating The Effect of
Financial Aid Offers on College Enrollment:A
Regression-Discontinuity Approach. International
Economic Review, 1249-1287.
M. H., S. S., & G. R. (2006). Evaluation of Knowledge
Management Technologies For the Support of
Technology Forecasting. 39th Annual Hawaii
Internationall Conference on System Sciences (pp. 1-
10). Systems Sciences.
M. S., & C. N. (2010). Enrolling in Higher Education:
The Perception of Stakeholders. Journal of
Institutional Research, 9-15.
M. S., & K. D. (2005). A New Trend Heuristic Time-
Variant Fuzzy Time Series Method for Forecasting
Enrollments. Computer and Information Sciences-
ISCIS, 553-564.
M. S., & K. D. (2005). Forecasting Enrollment Model
Based on First Order Fuzzy Time Series. In World
Academy of Science,Engineering and Technology,
375-378.
M. S., & M. H. (2012). Predicting Student Success from
Student Enrolment Data using Decision Tree
Technique. International Journal of Applied
Information Systems (IJAIS), 1-6.
Melike Seh, & K. D. (2005). Forecasting Enrollment
Model Based on First-Order Fuzzy Time Series.
Proceedings of World Academy of Science,
Engineering and Technology (pp. 375-378). WASET.
Ming, J. K. (2010). Institutional Factors Influencing
Students Choice Decision in Malaysia: A Conceptual
Framework. International Journal of Business and
Social Science, 53-58.
Ming, J. K. (2011). A Model of Higher Education
Institutions Choice in Malaysia - A Conceptual
Approach. International Proceedings of Economics
Development & Research.
Knowledge Management International Conference (KMICe) 2014, 12 – 15 August 2014, Malaysia
http://www.kmice.cms.net.my/ 662
Monks, J. (2009). The Impact of Merit Based Financial
Aid and Price Illusion on College Enrollment: A Field
Experiment. Elsevier, 99-106.
N. I., F. H., & N. M. (2013). International Students
Satisfaction Formation - Linkage Between
Information Satisfaction and Choice Satisfaction:
Study on Private Higher Education Institutions in
Malaysia. AFBE Journal, 88-108.
O. Z., M. J., & A. I. (2013). Factors Influencing Students'
Decisions in Choosing Private Institutions of Higher
Equations in malaysia: A Structural Equation
Modelling Approach. Asian Academy of Management,
1-16.
P. P., A. S., P. P., & S. N. (2008). Using Bayesian
Network for Planning Course Registration Model for
Undergraduate Students. Second IEEE International
Conference on Digital Ecosystems and Technologies
(IEEE DEST 2008), 492-496.
P. S., K. S., & S. E. (2012). Forecasting Enrollments
based on Fuzzy Time Series with Higher Forecast
Accuracy Rate. International Journal of Computer
Technology & Applications, 957-961.
Pathom Pumpuang, A. S., & P. P. (2008). Comparison of
Classifier Algorithms: Bayesian Network, C4.5,
Decision Forest and NBTree for Course registration
Planning Model of Undergraduate Students.
International Conference on Systems, Man and
Cybernetics, 3467-3651.
Q.Song, & B.S. Chissom. (1993). Fuzzy Time Series and
Its Models. Fuzzy Sets and Systems, Vol. 54, 269-277.
Quinn, M. (2013). Analyzing Long term University
Enrollment Dynamics using Monte Carlo Method.
AFBE journal.
R. L., & M. B. (2012). Enrollment Forecasting for School
Management System. International Journal of
Modelling and Optimization, 563-566.
Roper, A. T., S. C., A. P., T. M., F. R., & J. B. (2011).
Forecasting and Management of Technology. New
Jersey: John Wiley & Sons.
S. C., C. S., C. G., & W. W. (2007). Enrollment
Forecasting for an Upper Division General Education
Component. ASEE/IEEE Frontiers in Education
Conference, 25-28.
S. G. (2002). Three Enrollment Forecasting Models:
Issues in Enrollment projection for Community
Colleges. 40th RP Conference, (pp. 1-9). California.
S. M., S. H., & T. H. (2003). An Explatory Study of
Factors Influencing the College Choice Decision of
Undergraduate Students in Malaysia. Asia Pacific
Management Review, 259-280.
S.M.Chen. (2002). Forecasting Enrolments based on
High Order Fuzzy Time Series. Cybernetics and
Systems: An International Journal, 1-18.
Svetlana S. Aksenova, Du Zhang, & Meiliu Lu. (2006).
Enrollment Prediction through Data Mining. IEEE
International Conference on Information Reuse and
Integration, 510-515.
Sweeney, S. H., & Middleton, E. J. (2005). Multiregional
Cohort Enrollment Projections: Matching Methods to
Enrollment Policies. Population, Space and Place,
11(5), 361-380.
T. C. (n.d.). Data Mining: A Magic Technology for
College Recruitment. 1-18.
Vialardi, C., Chue, J., Peche, J., Alvarado, G., Estrella, J.,
& Ortigosa, A. (2011). A Data Mining Approach to
Guide Students through the Enrollment Process based
on Academic Performance. User Modeling and User
Adapted Interaction, 217-248.
Wagner, K., & P. Y. (2009). Factors Influencing
Malaysian Students' Intention to Study at a Higher
Educational Institution. E-Leader Kuala Lumpur, 1-
12.
Wong, W. K., E. B., & Chu, A. W. (2010). Adaptive
Time Variant Models for Fuzzy Time Series
Forecasting. IEEE Transaction on Systems, Man and
Cybernetics (pp. 1531-1541). IEEE.
Z. Ismail, & R.Efendi. (2011). Enrollment Forecasting
Based on Modified Weight Fuzzy Time Series.
Journal of Artificial Intelligence, 111-118.