=Paper=
{{Paper
|id=None
|storemode=property
|title=Empirical Investigation of Multi-tier Ensembles for the Detection of Cardiac Autonomic Neuropathy Using Subsets of the Ewing Features
|pdfUrl=https://ceur-ws.org/Vol-944/cihealth1.pdf
|volume=Vol-944
}}
==Empirical Investigation of Multi-tier Ensembles for the Detection of Cardiac Autonomic Neuropathy Using Subsets of the Ewing Features==
https://ceur-ws.org/Vol-944/cihealth1.pdf
Empirical Investigation of Multi-tier Ensembles
for the Detection of Cardiac Autonomic
Neuropathy Using Subsets of the Ewing Features
J. Abawajy1 , A.V. Kelarev1 , A. Stranieri2 , H.F. Jelinek3
1
School of Information Technology, Deakin University
221 Burwood Highway, Burwood, Victoria 3125, Australia
{jemal.abawajy,kelarev}@deakin.edu.au
2
School of Science, Information Technology and Engineering
University of Ballarat, P.O. Box 663, Ballarat, Victoria 3353, Australia
a.stranieri@ballarat.edu.au
3
Centre for Research in Complex Systems and School of Community Health
Charles Sturt University, P.O. Box 789, Albury, NSW 2640, Australia
hjelinek@csu.edu.au
Abstract. This article is devoted to an empirical investigation of per-
formance of several new large multi-tier ensembles for the detection of
cardiac autonomic neuropathy (CAN) in diabetes patients using sub-
sets of the Ewing features. We used new data collected by the diabetes
screening research initiative (DiScRi) project, which is more than ten
times larger than the data set originally used by Ewing in the investiga-
tion of CAN. The results show that new multi-tier ensembles achieved
better performance compared with the outcomes published in the litera-
ture previously. The best accuracy 97.74% of the detection of CAN has
been achieved by the novel multi-tier combination of AdaBoost and Bag-
ging, where AdaBoost is used at the top tier and Bagging is used at the
middle tier, for the set consisting of the following four Ewing features:
the deep breathing heart rate change, the Valsalva manoeuvre heart rate
change, the hand grip blood pressure change and the lying to standing
blood pressure change.
1 Introduction
Cardiac autonomic neuropathy (CAN) is a condition associated with damage
to the autonomic nervous system innervating the heart and highly prevalent in
people with diabetes, [6, 7, 24]. The detection of CAN is important for timely
treatment, which can lead to an improved well-being of the patients and a re-
duction in morbidity and mortality associated with cardiac disease in diabetes.
This article is devoted to empirical investigation of the performance of novel
large binary multi-tier ensembles in a new application for the detection of cardiac
autonomic neuropathy (CAN) in diabetes patients using subsets of the Ewing
1
features. This new construction belongs to the well known general and productive
multi-tier approach, considered by the first author in [14, 15].
Standard ensemble classifiers can generate large collections of base classi-
fiers, train them and combine into a common classification system. Here we deal
with new large multi-tier ensembles, combining diverse ensemble techniques on
two tiers into one scheme, as illustrated in Figure 1. Arrows in the diagram
correspond to the generation and training stage of the system, and show that
tier 2 ensemble generates and trains tier 1 ensembles and executes them in the
same way as it is designed to handle simple base classifiers. In turn, each tier 1
ensemble applies its method to the base classifier in the bottom tier.
Tier 2 Ensemble
Tier 1 Tier 1 Tier 1
Ensemble Ensemble Ensemble
Base Base Base Base
Classifier Classifier Classifier Classifier
Fig. 1. The generation and training stage of multi-tier ensembles
Large multi-tier ensembles illustrated in Figure 1 have not been considered
in the literature before in this form. They can be also regarded as a contribution
to the very large and general direction of research devoted to the investigation
of various multi-stage and multi-step approaches considered previously by other
authors. Let us refer to [1, 14, 15] for examples, discussion and further references.
Our experiments used the Diabetes Screening Complications Research Initia-
tive (DiScRi) data set collected at Charles Sturt University, Albury, Australia.
DiScRi is a very large and unique data set containing a comprehensive collection
of tests related to CAN. It has previously been considered in [5, 13, 21–23],
For the large DiScRi data set our new multi-tier ensembles produced better
outcomes compared with those published in the literature previously. Our new
results using multi-tier ensembles achieved substantially higher accuracies.
The paper is organised as follows. Section 2 describes the Diabetes Com-
plications Screening Research Initiative, cardiac autonomic neuropathy and the
2
Ewing features. Section 3 deals with the base classifiers and standard ensemble
classifiers. Section 4 describes our experiments and presents the experimental
results comparing the effectiveness of base classifiers, ensemble classifiers and
multi-tier ensembles for several subsets of the Ewing features. These outcomes
are discussed in Section 5. Main conclusions are presented in Section 6.
2 Diabetes Complications Screening Research Initiative
and the Ewing Features
This paper analysed the data set of test results and health-related parameters
collected at the Diabetes Complications Screening Research Initiative, DiScRi,
organised at Charles Sturt University, [5]. The collection and analysis of data has
been approved by the Ethics in Human Research Committee of the university
before investigations started. People participating in the project were attracted
via advertisements in the media. The participants were instructed not to smoke
and refrain from consuming caffeine containing drinks and alcohol for 24 hours
preceding the tests as well as to fast from midnight of the previous day until
tests were complete. The measurements were recorded in the DiScRi data base
along with various other health background data including age, sex and diabetes
status, blood pressure (BP), body-mass index (BMI), blood glucose level (BGL),
and cholesterol profile. Reported incidents of a heart attack, atrial fibrillation
and palpitations were also recorded.
The most essential tests required for the detection of CAN rely on assessing
responses in heart rate and blood pressure to various activities, usually consisting
of five tests described in [6] and [7]. Blood pressure and heart rate are very
important features [2], [29]. The most important set of features recorded for
detection of CAN is the Ewing battery [6], [7]. There are five Ewing tests in the
battery: changes in heart rate associated with lying to standing, deep breathing
and valsalva manoeuvre and changes in blood pressure associated with hand grip
and lying to standing. In addition features from ten second samples of 12-lead
ECG recordings for all participants were extracted from the data base. These
included the QRS, PQ, QTc and QTd intervals, heart rate and QRS axis. (QRS
width has also been shown to be indicative of CAN [9] and is included here.)
It is often difficult for clinicians to collect all test data. Patients are likely
to suffer from other illnesses such as respiratory or cardiovascular dysfunction,
obesity or arthritis, making it hard to follow correct procedures for all tests.
This is one of the reasons why it particularly important to investigate various
subsets of the Ewing battery.
The QRS complex and duration reflects the depolarization of the ventricles
of the heart. The time from the beginning of the P wave until the start of the
next QRS complex is the PQ interval. The period from the beginning of the QRS
complex to the end of the T wave is denoted by QT interval, which if corrected
for heart rate becomes the QTc. It represents the so-called refractory period of
the heart. The difference of the maximum QT interval and the minimum QT
interval over all 12 leads represents the QT dispersion (QTd). It is used as an
3
indicator of the repolarisation of the ventricles. The deflection of the electrical
axis of the heart measured in degrees to the right or left is called the QRS axis.
The whole DiScRi database contains over 200 features. We used the following
notation for the Ewing features and the QRS width:
LSHR stands for the lying to standing heart rate change;
DBHR is the deep breathing heart rate change;
V AHR is the Valsalva manoeuvre heart rate change;
HGBP is the hand grip blood pressure change;
LSBP is the lying to standing blood pressure change;
QRS is the width of the QRS segment, which is also known as a highly
significant indicator of CAN [9].
The detection of CAN deals with a binary classification where all patients
are divided into one of two classes: a ‘normal’ class consisting of patients without
CAN, and a ‘definite’ class of patients with CAN. Detection of CAN allows clini-
cians to collect fewer tests and can be performed with higher accuracy compared
with multi-class classifications of CAN progression following more detailed defi-
nitions of CAN progression classes originally introduced by Ewing. More details
on various tests for CAN are given in the next section. This paper is devoted to
the detection of CAN using subsets of the Ewing features.
A preprocessing system was implemented in Python to automate several ex-
pert editing rules that can be used to reduce the number of missing values in
the database. These rules were collected during discussions with the experts
maintaining the database. Most of them fill in missing entries of slowly chang-
ing conditions, like diabetes, on the basis of previous values of these attributes.
Preprocessing of data using these rules produced 1299 complete rows with com-
plete values of all fields, which were used for the experimental evaluation of the
performance of data mining algorithms.
3 Binary Base Classifiers and Standard Ensemble
Methods
Initially, we ran preliminary tests for many binary base classifiers available in
Weka [12] and included the following classifiers for a series of complete tests with
outcomes presented in Section 4. These robust classifiers were chosen since they
represent most essential types of classifiers available in Weka [12] and performed
well for our data set in our initial preliminary testing:
• ADTree classifier trains an Alternating Decision Tree, as described in [10].
Weka implementation of ADTree could process only binary classes.
• J48 generates a pruned or unpruned C4.5 decision tree [31].
• LibSVM is a library for Support Vector Machines [8]. It can handle only
attributes without missing values and only binary classes.
• NBTree uses a decision tree with naive Bayes classifiers at the leaves, [28].
• RandomForest constructs a forest of random trees following [4].
4
• SMO uses Sequential Minimal Optimization for training a support vector
classifier, [19, 30]. Initially, we tested all kernels of SMO available in Weka
and used it with polynomial kernel that performed best for our data set.
We used SimpleCLI command line in Weka [12] to investigate the perfor-
mance of the following ensemble techniques:
• AdaBoost training every successive classifier on the instances that turned
out more difficult for the preceding classifier [11];
• Bagging generating bootstrap samples to train classifiers and amalgamating
them via a majority vote, [3];
• Dagging dividing the training set into a disjoint stratified samples [33];
• Grading labelling base classifiers as correct or wrong [32];
• MultiBoosting extending AdaBoost with the wagging [34];
• Stacking can be regarded as a generalization of voting, where meta-learner
aggregates the outputs of several base classifiers, [35].
We used SimpleCLI command line in Weka [12] to train and test multi-tier
ensembles of binary classifiers too.
4 Experimental Results
We used 10-trial 10-fold cross validation to evaluate the effectiveness of classifiers
in all experiments. It is often difficult to obtain results for all five tests and we
therefore included the largest subsets of four features from the Ewing battery.
These subsets can help clinicians to determine whether CAN is present in those
situations when one of the tests is missing. The following notation is used to
indicate these subsets in the tables with outcomes of our experiments:
SEwing is the set of all five Ewing features, i.e., LSHR, DBHR, V AHR,
HGBP and LSBP ;
SLSHR is the set of four Ewing features with LSHR excluded, i.e.,
DBHR, V AHR, HGBP and LSBP ;
SDBHR is the set of four Ewing features with DBHR excluded, i.e.,
LSHR, V AHR, HGBP and LSBP ;
SV AHR is the set of four Ewing features with V AHR excluded, i.e.,
LSHR, DBHR, HGBP and LSBP ;
SHGBP is the set of four Ewing features with HGBP excluded, i.e.,
LSHR, DBHR, V AHR and LSBP ;
SLSBP is the set of four Ewing features with LSBP excluded, i.e.,
LSHR, DBHR, V AHR and HGBP ;
S4 is the set of two heart rate features LSHR, DBHR, one
blood pressure feature HGBP , with QRS added.
Feature selection methods are very important, see [25], [26], [27]. In particular,
the set S4 was identified by the authors in [13] using feature selection.
5
First, we compared the effectiveness of base classifiers for these sets of fea-
tures. We used accuracy to compare the classifiers, since it is a standard measure
of performance. The accuracy of a classifier is the percentage of all patients clas-
sified correctly. It can be expressed as the probability that a prediction of the
classifier for an individual patient is correct. The experimental results compar-
ing all base classifiers are included in Table 1. These outcomes show that for the
DiScRi database RandomForest is the most effective classifier. It is interesting
that many classifiers worked more accurately when the LSHR feature had been
excluded.
Subsets of features
SEwing SLSHR SDBHR SV AHR SHGBP SLSBP S4
ADTree 84.14 84.68 75.31 80.08 81.02 71.73 80.77
J48 91.61 92.15 85.14 90.92 91.28 89.99 91.38
LibSVM 92.39 92.94 80.97 92.71 85.82 84.78 91.13
NBTree 90.15 91.07 81.83 87.45 87.22 86.99 87.76
RandomForest 94.46 94.84 91.76 93.61 94.23 93.76 94.35
SMO 74.13 73.75 64.36 71.98 73.83 71.36 74.44
Table 1. Accuracy of base classifiers for the detection of CAN using subsets of Ewing
features
Second, we compared several ensemble classifiers in their ability to improve
the results. Preliminary tests demonstrated that ensemble classifiers based on
RandomForest were also more effective than the ensembles based on other clas-
sifiers. We compared AdaBoost, Bagging, Dagging, Grading, MultiBoost and
Stacking based on RandomForest. The accuracies of the resulting ensemble clas-
sifiers are presented in Table 2, which shows improvement. We used one and the
same base classifier, RandomForest, in all tests included in this table. We tested
several other ensembles with different base classifiers, and they turned out worse.
Finally, we compared the results obtained by all multi-tier ensembles combin-
ing AdaBoost, Bagging and MultiBoost, since these ensembles produced better
accuracies in Table 2. Tier 2 ensemble treats the tier 1 ensemble and executes
it in exactly the same way as it handles a base classifier. In turn the tier 1
ensemble applies its method to the base classifier as usual. We do not include
repetitions of the same ensemble technique in both tiers, since such repetitions
were less effective. The outcomes of the multi-tier ensembles of binary classifiers
are collected in Tables 3.
6
Subsets of features
SEwing SLSHR SDBHR SV AHR SHGBP SLSBP S4
AdaBoost 96.84 97.23 94.07 95.99 96.59 96.11 96.51
Bagging 96.37 96.75 93.63 95.52 96.13 95.67 96.05
Dagging 89.75 90.13 87.18 88.94 89.54 89.10 89.46
Grading 94.49 94.87 91.79 93.61 94.26 93.78 94.18
MultiBoost 96.37 96.77 93.62 95.50 96.13 95.65 96.04
Stacking 95.44 95.81 92.70 94.56 95.20 94.73 95.09
Table 2. Accuracy of ensemble classifiers for the detection of CAN using subsets of
Ewing features
Subsets of features
Tier 2 Tier 1 SEwing SLSHR SDBHR SV AHR SHGBP SLSBP S4
AdaBoost Bagging 97.35 97.74 94.58 96.50 97.12 96.65 97.04
AdaBoost MultiBoost 96.37 96.78 93.65 95.52 96.14 95.68 96.07
Bagging AdaBoost 97.33 97.73 94.57 96.49 97.09 96.61 97.00
Bagging MultiBoost 96.66 97.08 93.91 95.80 96.42 95.96 96.34
MultiBoost AdaBoost 96.85 97.25 94.08 96.00 96.62 96.13 96.52
MultiBoost Bagging 97.05 97.43 94.30 96.19 96.83 96.35 96.74
Table 3. Accuracy of multi-tier ensembles of binary classifiers for the detection of
CAN using subsets of the Ewing features
7
5 Discussion
DiScRi is a very large and unique data set containing a comprehensive collection
of tests related to CAN. It has been previously considered in [13, 21–23], New
results obtained in this paper achieved substantially higher accuracies than the
previous outcomes published in [13]. Overall, the results of the present paper are
also appropriate for other data mining applications in general when compared
to recent outcomes obtained for other data sets using different methods, for
example, in [16] and [17].
AdaBoost has produced better outcomes than other ensemble methods for
subsets of the Ewing features of the DiScRi data set; and the best outcomes
were obtained by a novel combined ensemble classifier where AdaBoost is used
after Bagging.
There are several reasons, why other techniques turned out less effective.
First, Dagging uses disjoint stratified training sets to create an ensemble, which
benefits mainly classifiers of high complexity. Our outcomes demonstrate that
the base classifiers considered in this paper are fast enough and this benefit
was not essential. Second, stacking and grading use an ensemble classifier to
combine the outcomes of base classifiers. These methods are best applied to
combine diverse collections of base classifiers. In this setting stacking performed
worse than bagging and boosting.
Our experiments show that such large multi-tier ensembles of binary classi-
fiers are in fact fairly easy to use and can also be applied to improve classifi-
cations, if diverse ensembles are combined at different tiers. It is an interesting
question for future research to investigate multi-tier ensembles for other large
datasets.
6 Conclusion
We have investigated the performance of novel multi-tier ensembles for the de-
tection of cardiac autonomic neuropathy (CAN) using subsets of the Ewing fea-
tures. Our experimental results show that large multi-tier ensembles can be used
to increase the accuracy of classifications. They have produced better outcomes
compared with previous results published in the literature. The best accuracy
97.74% of the detection of CAN has been achieved by the novel multi-tier com-
bination of AdaBoost and Bagging, where AdaBoost is used at the top tier and
Bagging is used at the middle tier, for the set consisting of the following four
Ewing features: the deep breathing heart rate change, the Valsalva manoeuvre
heart rate change, the hand grip blood pressure change and the lying to standing
blood pressure change. This level of accuracy is also quite good in comparison
with the outcomes obtained recently for other data sets in closely related areas
using different methods, for example, in [18, 20, 16, 17, 36].
8
Acknowledgements
This work was supported by a Deakin-Ballarat collaboration grant. The authors
are grateful to four referees for comments that have helped to improve the pre-
sentation, and for suggesting several possible directions for future research.
References
1. Al-Ani, A., Deriche, M.: A new technique for combining classifiers using the
Dempster-Shafer theory of evidence. Journal of Artificial Intelligence Research 17,
333–361 (2012)
2. Al-Jaafreh, M., Al-Jumaily, A.A.: New model to estimate mean blood pressure
by heart rate with stroke volume changing influence. In: IEEE EMBC 2006: 28th
Annual International Conference of the IEEE Engineering in Medicine and Biology
Society. pp. 1803–1805 (2006)
3. Breiman, L.: Bagging predictors. Machine Learning 24, 123–140 (1996)
4. Breiman, L.: Random Forests. Machine Learning 45, 5–32 (2001)
5. Cornforth, D., Jelinek, H.F.: Automated classification reveals morphological factors
associated with dementia. Applied Soft Computing 8, 182–190 (2007)
6. Ewing, D.J., Campbell, J.W., Clarke, B.F.: The natural history of diabetic auto-
nomic neuropathy. Q. J. Med. 49, 95–100 (1980)
7. Ewing, D.J., Martyn, C.N., Young, R.J., Clarke, B.F.: The value of cardiovascular
autonomic function tests: 10 years experience in diabetes. Diabetes Care 8, 491–498
(1985)
8. Fan, R.E., Chen, P.H., Lin, C.J.: Working set selection using second order infor-
mation for training SVM. J. Machine Learning Research 6, 1889–1918 (2005)
9. Fang, Z.Y., Prins, J.B., Marwick, T.H.: Diabetic cardiomyopathy: evidence, mech-
anisms, and therapeutic implications. Endocr. Rev. 25, 543–567 (2004)
10. Freund, Y., Mason, L.: The alternating decision tree learning algorithm. In: Proc.
16th Internat. Conf. Machine Learning. pp. 124–133 (1999)
11. Freund, Y., Schapire, R.E.: Experiments with a new boosting algorithm. In: Proc.
13th Internat. Conf. Machine Learning. pp. 148–156 (1996)
12. Hall, M., Frank, E., Holmes, G., Pfahringer, B., Reutemann, P., Witten, I.H.: The
WEKA data mining software: an update. SIGKDD Explorations 11, 10–18 (2009)
13. Huda, S., Jelinek, H.F., Ray, B., Stranieri, A., Yearwood, J.: Exploring novel fea-
tures and decision rules to identify cardiovascular autonomic neuropathy using a
hybrid of wrapper-filter based feature selection. In: Sixth International Conference
on Intelligent Sensors, Sensor Networks and Information Processing, ISSNIP 2010.
pp. 297–302. IEEE Press, Sydney (2010)
14. Islam, R., Abawajy, J.: A multi-tier phishing detection and filtering approach.
Journal of Network and Computer Applications p. to apper soon (2012)
15. Islam, R., Abawajy, J., Warren, M.: Multi-tier phishing email classification with an
impact of classifier rescheduling. In: 10th International Symposium on Pervasive
Systems, Algorithms, and Networks, ISPAN 2009. pp. 789–793 (2009)
16. Jelinek, H.F., Khandoker, A., Palaniswami, M., McDonald, S.: Heart rate variabil-
ity and QT dispersion in a cohort of diabetes patients. Computing in Cardiology
37, 613–616 (2010)
9
17. Jelinek, H.F., Rocha, A., Carvalho, T., Goldenstein, S., Wainer, J.: Machine learn-
ing and pattern classification in identification of indigenous retinal pathology. In:
33rd Annual International Conference of the IEEE Engineering in Medicine and
Biology Society. pp. 5951–5954. IEEE Press (2011)
18. Kang, B., Kelarev, A., Sale, A., Williams, R.: A new model for classifying DNA
code inspired by neural networks and FSA. In: Advances in Knowledge Acquisi-
tion and Management. Lecture Notes in Computer Science, vol. 4303, pp. 187–198
(2006)
19. Keerthi, S.S., Shevade, S.K., Bhattacharyya, C., Murthy, K.R.K.: Improvements
to Platt’s SMO algorithm for SVM classifier design. Neural Computation 13(3),
637–649 (2001)
20. Kelarev, A., Kang, B., Steane, D.: Clustering algorithms for ITS sequence data with
alignment metrics. In: AI 2006: Advances in Artificial Intelligence, 19th Australian
Joint Conference on Artificial Intelligence. Lecture Notes in Artificial Intelligence,
vol. 4304, pp. 1027–1031 (2006)
21. Kelarev, A.V., Dazeley, R., Stranieri, A., Yearwood, J.L., Jelinek, H.F.: Detection
of CAN by ensemble classifiers based on Ripple Down Rules. In: Pacific Rim Knowl-
edge Acquisition Workshop, PKAW2012. Lecture Notes in Artificial Intelligence,
vol. 7457, pp. 147–159 (2012)
22. Kelarev, A.V., Stranieri, A., Yearwood, J.L., Jelinek, H.F.: Empirical investigation
of consensus clustering for large ECG data sets. In: 25th International Symposium
on Computer Based Medical Systems, CBMS2012. pp. 1–4 (2012)
23. Kelarev, A.V., Stranieri, A., Yearwood, J.L., Jelinek, H.F.: Empirical study of
decision trees and ensemble classifiers for monitoring of diabetes patients in perva-
sive healthcare. In: Network-Based Information Systems, NBiS-2012. pp. 441–446
(2012)
24. Khandoker, A.H., Jelinek, H.F., Palaniswami, M.: Identifying diabetic pa-
tients with cardiac autonomic neuropathy by heart rate complexity anal-
ysis. BioMedical Engineering OnLine 8, http://www.biomedical–engineering–
online.com/content/8/1/3 (2009)
25. Khushaba, R., Al-Jumaily, A., Al-Ani, A.: Evolutionary fuzzy discriminant analysis
feature projection technique in myoelectric control. Pattern Recognition Letters 30,
699–707 (2009)
26. Khushaba, R., AlSukker, A., Al-Ani, A., Al-Jumaily, A.: A novel swarm-based fea-
ture selection algorithm in multifunction myoelectric control. Journal of Intelligent
& Fuzzy Systems 20, 175–185 (2009)
27. Khushaba, R.N., Al-Ani, A., Al-Sukker, A., Al-Jumaily, A.: A combined ant colony
and differential evolution feature selection algorithm. In: ANTS2008: Ant Colony
Optimization and Swarm Intelligence. Lecture Notes in Computer Science, vol.
5217, pp. 1–12 (2008)
28. Kohavi, R.: Scaling up the accuracy of Naive-Bayes classifiers: a Decision-Tree
hybrid. In: Second International Conference on Knowledge Discovery and Data
Mining. pp. 202–207 (1996)
29. Mahmood, U., Al-Jumaily, A., Al-Jaafreh, M.: Type-2 fuzzy classification of blood
pressure parameters. In: ISSNIP2007: The third International Conference on Intel-
ligent Sensors, Sensor Networks and Information Processing. pp. 595–600 (2007)
30. Platt, J.: Fast training of support vector machines using sequential minimal opti-
mization. In: Advances in Kernel Methods – Support Vector Learning (1998)
31. Quinlan, R.: C4.5: Programs for Machine Learning. Morgan Kaufmann, San Mateo,
CA (1993)
10
32. Seewald, A.K., Fuernkranz, J.: An evaluation of grading classifiers. In: Advances
in Intelligent Data Analysis. Lecture Notes in Computer Science, vol. 2189/2001,
pp. 115–124 (2001)
33. Ting, K.M., Witten, I.H.: Stacking bagged and dagged models. In: Fourteenth
International Conference on Machine Learning. pp. 367–375 (1997)
34. Webb, G.I.: MultiBoosting: a technique for combining boosting and wagging. Ma-
chine Learning 40, 159 – 196 (2000)
35. Wolpert, D.H.: Stacked generalization. Neural Networks 5, 241–259 (1992)
36. Yearwood, J.L., Kang, B.H., Kelarev, A.V.: Experimental investigation of classi-
fication algorithms for ITS dataset. In: Pacific Rim Knowledge Acquisition Work-
shop, PKAW 2008. pp. 262–272. Hanoi, Vietnam, 15–16 December 2008 (2008)
11