Sie sind auf Seite 1von 8

IEEE-201S0

MULTIPLE CLASSIFICATION SYSTEM


FOR FRACTURE DETECTION IN
HUMAN BONE X-RAY IMAGES
# *
N.Umadevi , Dr.S.N.GeethaJakshmi
#HOD of Computer Science & IT, *Associate Professor
# *
Sri Jayendra Saraswathy Maha VidyaJaya College of Arts & Science, AvinashiJingam Deemed
University for Women
# *
Singanallur Coimbatore 641015 India, Coimbatore 641043 India
# umadevigayathri@rediffmail.com, sngeethalakshrni@yaoo.com
*

Abstract - X-Ray is one the oldest and frequently used presentation and their outcomes are unpredictable. Tibia
devices, that makes images of any bone in the body, fractures are the most common long bone fracture accounting
including the hand, wrist, arm, elbow, shoulder, foot, to more than 20 per cent occupancy of hospital wards [9].
ankle, leg (shin), knee, thigh, hip, pelvis or spine. A According to [27] , on average 26 tibia fractures occur per
typical bone ailment is the fracture, which occurs when 1,00,000 populations per year.
bone cannot withstand outside force like direct blows,
twisting injuries and falls. Automatic detection of The numerous incidences necessitate the healthcare
fractures in bone x-ray images is considered important, as professionals to analyze huge number of x-ray images. As a
humans are prone to miss-diagnosis. The main focus of consequence of continuously analyzing images, cases of miss
this paper is to automatically detect fractures in long diagnosis occur. A miss is defmed as either failure to see a
bones and in particular, leg bone (often referred as significant fmding, or attaching the incorrect significance to a
Tibia), from plain diagnostic X-rays using a multiple fmding that is readily seen. A high miss-diagnosis rate will
classification system. Two types of features (texture and result in poor quality in healthcare and time-delayed
shape) with three types of classifiers (Back Propagation treatment and has to be avoided. All these facts have
Neural Network, K-Nearest Neightbour, Support Vector motivated us to focus on the proposed fracture identification
Machine) are used during the design of multiple and detection system that can automatically determine the
classifiers. A total of 12 ensemble models are proposed. presence and absence of fractures and if present, methods that
Experiments proved that ensemble models significantly could locate the fractures in a bone x-ray image.
improve the quality of fracture identification. Classification, a frequently used data mining technique, has
been used widely to detect the presence of fracture for the
I. INTRODUCTION past few decades [2S]. These systems combine various
features (like shape, texture and colour) extracted from x-ray
In today's revolution oriented medical environment, images and machine learning algorithms to identify faults.
computer aided disease detection play a vital role in a wide
range of applications and services in day-to-day activities. According to [32] when a perfect set of features that can
The most important quality goal of such a system is speed describe the image data is given, the accuracy of the resultant
and high accuracy in disease detection. This is more classification depends on the classifier adopted. Several
important in radiology department, where hundreds and solutions have been proposed for this purpose. Among which,
hundreds of images are to be examined for various ailments the usage of Neural Network (NN), K Nearest Neighbour
including bone fracture, infection and location of foreign (KNN) and Support Vector Machines (SVM) based
objects. Out of this, bone fracture detection is the most classifiers are more prominent. However, as pointed out by
frequently used task, because the population affected by this [30], the success rate of a classification problem can be
ailment has increased due to the increased reports of improved by using multiple classifiers and then the various
accidents and due to osteoporosis. According to the market results can be fused together to obtain an optimum
report of [ IS], the fracture features are more in Indian classification. Using multiple classifiers (either different types
hospital records and the incidence has increased 3-folds over of classifiers or different instantiations of the same classifier)
the past 35 decades with more than 4.4 lakh people and is improve the success rate of an classification model. This
expected to increase to more than 6 lakh in 2020 [21]. process is termed as 'ensemble classification' or 'fusion
Among fractures, automatic detection in tibia is considered classification' or 'multiple classification system'. According
more chaUenging because they are different and variable in to [31], intuitively, fusion classification allows the different

ICCCNT'12
26 th_2Sth July 2012, Coimbatore, India
IEEE-20180

needs of a difficult problem to be handled by classifiers


suited to those particular needs. Mathematically, fusion
classifier provide an extra degree of freedom in the classical
bias/variance tradeoff, allowing solutions that would be
difficult (if not impossible) to reach with only a single
classifier. A general ensemble classification system when
applied to classifY x-ray images is shown in Figure I.

Multiclassifier Classification Result

Classifier 1
Fracture Present
Features
X-Ray Image Classifier 2
Fracture Absent
Classifier 3
L ________________________

Figure 1 : Classification Ensemble Model

The accuracy of such an ensemble model depends on several


factors like (i) Number and type of classifiers (ii) Features
Fusion classifiers or multiple classifier systems (MCS) have
used (iii) Partitioning method (Training and Testing sets) (iv)
received considerable attention in applied statistics [17],
The aggregation method and (v) Type of training. A
machine learning [13] and pattern recognition [24] for over a
successful ensembling depends on the design of the
decad. Several studies demonstrate that the practice of
individual classifiers, selection of classifier and design of the .
combmmg several base classifier models into one aggregated
fusion mechanism. Studies reported have shown that
classifier leads to significant gains in classification performance
ensembling classification is effective only when the
over its constituent members. Over the years, different fusion
individual classifiers selected exhibit two important
algorithms have been proposed, which differ along three
criterions, namely, high accuracy (or low error rate) and
structural dimensions of fusion design as listed below.
diversity between classifiers (make different error rate) [23].
Techniques to increase diversity include methods as listed
(i) the choice of the base or member class ifier
below.
(ii) the treatment of the input training data and
(iii) the aggregation strategy for the outputs of
1. Us ing different types of classifiers
member classifiers.
2. Training individual classifiers with different
data set (bagging, boosting and random
Firstly, two broad strategies exist for choosing the members
sequence)
of a fusion classifier [8]. In hybrid multiple classifier
3. Using different subsets of features
different types of algorithms are combined, whilst in non
4. Fusion of classification result
hybrid multiple classifiers, one classifier algorithm is chosen
as base classifier and replicated multiple times in order to
In the present study, for the first step three classifiers namely,
constitute a fusion.
Back Propagation Neural Network (BPNN), K-Nearest
Neighbour (KNN) and Support Vector Machine (SVM) are
Secondly, many algorithms differ in terms of the treatment of
used. For the second step, techniques like bagging, boosting
the training data, used as input for each base classifier.
and random sequence methods are used. The feature subsets
Possibilities include data sampling schemes [4], variable
are created using lO-fold cross validation method. The
selection [19] or more complex data transformations [25],
weighted majority voting rules is used during fusion process.
[34]. A third design characteristic involves the fusion rule
The rest of the paper is organized as follows. Section 2
used for the fusion of member outputs, ranging from simple
discusses the various features and classifiers used in this
average aggregation to more complex combination rules
study, while Section 3 presents techniques for the second
third and fourth steps. Section 4 presents the experimental
[36].
results while Section 5 concludes the work with future
research directions. The most popular multiple classifier schemes are non-hybrid,
where a base classification algorithm is applied to differently
permutated training sets. A well-known method in this
II. EXISTING METHODS
category is Bagging [5], an acronym of bootstrap aggregating.
Although numerous variations have been proposed since its
This section presents reviews of the various methods
introduction [11], [20]. Breiman's original implementation is
that exist in the area of multiple classification problem.
still a widely used multiple classifier. In Bagging, each
ember is trained on a bootstrap sample of the training data,
l.e. a random sample of observations drawn with replacement

ICCCNT'12
26 th_28th July 2012, Coimbatore, India
IEEE-20 ISO

and having the same size as the original training data. The preprocessing step enhances the x-ray images through the
Multiple classification is obtained by means of uniform use of a hybrid denoising method that uses Independent
majority voting, where an unlabeled observation is assigned Component Analysis (lCA) coupled with wavelets. The
the class with the highest number of votes among the implementation details are given in [37]. The second step,
individual classifiers' predictions. Theoretically, segmentation is used to extract the bone structure from the x
bootstrapping can induce large differences in the constructed ray images. For this purpose, an enhanced active contour
individual classifiers which substantially improve the model that uses region growing algorithm to estimate the
accuracy of the fusion classifier. Several variations upon initial seeds is used. Further, as the region of interest to detect
Bagging have been proposed in search for further fractures is considered as the diaphysis of tibia, a
performance improvements. Two popular strategies involve segmentation method that extracts only the diaphysis region
(i) increasing variation in the training data for base classifiers is proposed. This method uses a fast hough transformation
and (ii) the use of alternative base classifier algorithms. method. Detailed description of the segmentation process is
Firstly, several studies have shown the impact of variations presented in [39]. General details on tibia and diaphysis can
of the input data used for the training of base classifiers. be found in previous published in [3S].
Varying the training data of the members of an fusion
classifier is a strategy to increase diversity amongst member The feature extraction stage is used to extract various features
class ifiers, which is generally perceived as a key driver of that best exhibit the characteristics of the segmented image.
fusion performance [26]. In the Random Subspace Method Twelve features, which can be grouped into texture features
(RSM), variables are randomly sampled to create training and shape features, are used to generate the feature vector.
data sets for a decision tree classification fusion. RSM, also The texture features collected are GLCM (Gray Level Co
referred to as Attribute Bagging [6], specifies that each Occurrence Matrix) features, namely, Contrast, Homogeneity,
fusion member is trained using a random feature subset Energy, Entropy, Mean, Variance, Standard Deviation,
(RFS), i.e. a random selection of explanatory variables Correlation, Gabor orientation (GO), Markov Random Field
sampled without replacement and of a predefmed size. (MRF), and intensity gradient direction (IGD). The shape
features are extracted using a Fast Hough Transformation
A related method is the Random Forest algorithm by [3], proposed by [16]. The features collected are arranged in a
which has demonstrated high classification performance in two-dimensional matrix where each column represents a
many fields of research [1]. A Random Forest combines feature extracted and the row represents various features that
Bagging and a specific form of RSM where random feature represent the segmented x-ray image. The final column is
subset selection is performed at each node of a member treated as a target label column, having a Boolean value of 0
decision tree. Rotation Forest [34], a multiple classifier based to indicate the absence of fracture and 1 to indicate the
on rotations of the feature space through Principal presence of fracture. Us ing these 12 features, three feature set
Component Analysis (PCA). The purpose of Rotation Forest vectors are created. The frrst one consists of only texture
is to increase the individual classifier performance and the features, second have only shape features and the third set
diversity within the multiple classification process. Diversity have a shape and texture features and are referred to as FS I,
is achieved for each classifier by applying feature extraction, FS2 and FS I2 respectively in this paper.
while one tries to increase the performance by using all
principal components and training the model on the whole The present study considers three classifiers namely, Support
data set. Vector Machine (SVM) [15], K-Nearest Neighbour (KNN)
[10] and Back Propagation Artificial Neural Network
A second strategy to increase classification performance is to (BPNN) [7]. All the classifiers are designed as binary
select an alternative base classifier algorithm. Many studies classifiers, where the classification result is either 'Fracture
have proposed fusion based on alternative base classifiers, Present' or 'Fracture Absent'. During experimentation, a
such as Artificial Neural Networks [40], Support Vector standard three-layered back-propagation network with the
Machines [22], parametric regression techniques [33] and tangent-sigmoid transfer function is considered. The weights
nonparametric regression techniques [2]. and biases of the neural networks are initialized randomly,
and the number of neurons in the hidden node is determined
III. PROPOSED CLASSIFICATION SYSTEM heuristically as inputs + outputs. A small value of the learning
rate (0.15) and a large value of the momentum rate (O.S) are
The proposed ensemble classification system consists of six chosen to avoid local minima. The number of training epochs
steps. was 500. To implement the principles of SVMs, we used the
LIB-SVM. The two most important steps in implementation
l. Preprocessing of SVM is scaling and kernel selection; for scaling, the values
2. Segmentation of all features were linearly scaled to the range [1, + 1] to
3. Feature Extraction and Normalization prevent the cases that features great numeric ranges
4. Training of classifiers dominating those in smaller numeric ranges. Among many
5. Assess classifiers with respect to diversity available kernel functions linear kernel was used. The KNN
6. Perform class ification fusion algorithm was constructed using k=2.

ICCCNT'12
26 th_2Sth July 2012, Coimbatore, India
IEEE-20 IS0

The classifiers are trained using boosting [14] method where The fusion method used was majority voting method.
different feature subsets are used to train the classifiers. The Majority voting scheme is one of the oldest strategies used to
accuracy of the single classifiers is evaluated using lO-fold aggregate the results of multiple classifiers in an ensemble
cross validation technique. Table I shows the performance of system. It is one of the simplest and fastest algorithm and the
the three classifiers with respect to classification accuracy. method is explained as follows. Let the decision of the i th
classifier be defmed as dt,j E {O, I}, t I, . . . , T and j I, . . . ,
= =

Table 1 : Ten-fold Validation Classification Accuracy C, where T is the number of classifiers and C is the number of
classes. If the ith classifier chooses class (OJ, then dt,j I and 0,
=

otherwise. In majority voting scheme, a class (OJ is chosen, if

T c T
(2)
L dt.J = max L dt.j
The results from the above table show that while considering t=l J=l t=l
accuracy the classifiers are roughly comparable. As the next
step, the diversity of the classifiers is considered and are With majority voting the fusion classifier makes the correct
estimated using Yule's Q method [23], where Q between two
decision if at least L TI2J + I classifiers choose the correct
classifiers is calculated using the confusion matrix (Figure 2)
label, where the floor function L J returns the largest integer
.

and Equation (1).


less than or equal to its argument.

= ad- bc (1)
Q IV. EXPERIMENTAL RESULTS
ad+bc
The experimental results of ensemble classification with
Classifier I Classifier I different base classifiers are presented in the following
(Correct) (Incorrect) sections. To analyze the performance of the ensemble systems
Classifier 2 A b the commonly used measures, namely Accuracy, Precision,
(Correct) Recall and F-Measure are used. All the measures are
Classifier 2 C d calculated from the confusion matrix (Figure 2). Accuracy is
(Incorrect) a metric that is used to measure the correct classification rate
and is defmed as the ratio of number of images correctly
Figure 2 : Confusion Matrix predicted to the total number of images (Equation 3). The
second measure, precision, also used to analyze the

{<
The resultant Q value ranges between -I to I and can be correctness of a classifier is the ratio of the number of images
interpreted as follows: corrected classified as having fracture to the total number of
images predicted as having fracture (Equation 4). The recall,
0 Uncorrelated ClassifielS otherwise known as defect detection rate, is the ratio of
number of images corrected classified as having fracture to
= 0 Statistically Independent ClassifielS the total number of images that actually have fracture
(Equation 5). F Measure an amalgamation method of
> 0 CorrelatedClassifielS
precision and recall presents the accuracy in a more precise
manner and can be calculated using Equation (6).
Table 2 shows the Q values obtained for the three selected
classifiers.
a+d (3)
Accuracy= - -:-- --,
Table 2 Q Values for the Selected Classifiers
-
: a+b+c+d
d (4)
Precislon=
.. -

BPNN SVM KNN b+d


BPNN 1.000 0.791 0.312 d (5)
Recall=- -
SVM 0.791 1.000 0.225 c+d
KNN 0.312 0.225 1.000 2* Precision* Recall (6)
F - Measure= -------

Precision+Recall
From the table, it could be seen that even though all
algorithms are moderately correlated, they exhibit high To analyze the effectiveness of the ensemble classifiers,
pairwise diversities to each other. Thus, it could be seen that statistical analysis of the results was conducted. 't' test
the selected base classifiers have high diversity and can be method proposed by [29] was used for this purpose. The NB
used to design ensemble system. The present study considers 't' test was performed at 95% confidence level (0.05 level) to
two-classifier and three-class ifier ensemble systems. The analyze the significant difference between the single
classifiers were built while varying the feature sets. Thus, a classifiers and ensemble classifiers. The NB-'t' test method
total of 12 unique combinations of classifiers using three was adopted because it is more suited for classifiers adapting
single classifiers and 3 feature sets were designed. lO-fold cross-validation method [12]. The traditional student

ICCCNT'12
26 th_2Sth July 2012, Coimbatore, India
IEEE-201S0

't' test, method produces more false significant differences


due to the dependencies that exists in the estimates.

Tables 3 to 5 show the proposed 2-classifer and 3-classifier


ensemble models based on Accuracy, Precision, Recall and
F-Measure while using different feature sets. To analyze the
advantage obtained by the proposed predictors the proposed
models are compared with their traditional single classifier
counterparts. In the tables, the single classifiers BPNN,
KNN and SVM are denoted as A, B and C respectively.
Ensemble classifiers BPNN + SVM, BPNN + KNN, SVM +
KNN and BPNN + SVM + KNN are denoted as 1, 2, 3 and 4
respectively. Further, 'SD' denotes the standard deviation and
the column 'Sig' denotes the status of significance. In the Sig
column, 'Yes' denotes that there is a significance
performance difference between single prediction model and
the corresponding ensemble prediction model, while a 'No'
represents insignificant performance. A '+' sign at the end
denotes that ensemble prediction model has outperformed the
corresponding single prediction model, while '- ' sign
denotes the opposite.

From the results it could be seen that the application of


ensembling concept to detect fractures in x-ray images has
improved. The performance of class ifiers while using both
texture and shape features show significant improvement in
classification performance. However, the statistical results of
precision with KNN classifier show that there is significant
difference between the ensemble models and their
corresponding single-classifier base models. But the
precision column of KNN classifier shows negative
insignificance indicating performance degradation when
compared with the single KNN classifier modeL But, the
recall parameter, which plays more important role in
classification, achieved positive significant difference. The
same models when compared with F measure also showed
significant positive difference and outperformed the base
modeL While comparing the three classifiers, the
performance of SVM-based prediction models is better when
compared with BPNN and KNN. While considering the
number of classifiers, the 3-classifier ensemble model ranked
first when compared with all other models. Thus, among the
12 proposed model, the best performance was produced by
the model that used fusion techniques that combines BPNN,
KNN and SVM classifiers with texture and shape features.

V.CONCLUSION

In this paper, the application of ensemble classification to


fracture detection in x-ray images was considered. Two types
of features, namely, texture features and shape features were
extracted from the x-ray images forming a total of 12
features. Three binary classifiers, SVM, BPNN and KNN
were to build ensemble classification models and during
training, boosting method was used. The present study
considers two-classifier and three-classifier ensemble
systems. The classifiers were built while varying the feature
sets.

ICCCNT'12
26 th_2Sth July 2012, Coimbatore, India
IEEE-201S0

Table 3: Performance of Ensemble Models using Texture feature set

Accuracy Precision Recall F-Measure


Model
Mean SD Sig Mean SD Sig Mean SD Sig Mean SD Sig
A 77.38 3.562 80.12 2.981 84.01 3.015 82.02 3.298
1 94.55 1.579 Yes(+) 98.93 0.371 Yes(+) 92.94 1.574 Yes(+) 95.84 0.361 Yes(+)
2 89.91 1.236 Yes(+) 97.36 0.899 Yes(+) 93.44 0.587 Yes(+) 95.36 0.745 Yes(+)
4 96.17 1.314 Yes(+) 99.94 0.012 Yes(+) 94.16 1.122 Yes(+) 96.96 0.202 Yes(+)
B 84.98 2.416 89.72 0.126 95.42 0.124 92.48 0.397
2 89.91 1.236 Yes(+) 97.36 0.899 Yes(+) 93.44 0.587 Yes(+) 95.36 0.745 Yes(+)
3 92.26 1.077 Yes(+) 97.94 0.821 Yes(+) 92.67 0.687 Yes(+) 95.23 0.798 Yes(+)
4 96.17 1.314 Yes(+) 99.94 0.012 Yes(+) 94.16 1.122 Yes(+) 96.96 0.202 Yes(+)
C 90.62 1.161 90.34 0.040 98.43 0.068 94.21 1.014
I 94.55 1.579 Yes(+) 98.93 0.371 Yes(+) 92.94 1.574 Yes(+) 95.84 0.361 Yes(+)
3 92.26 1.077 Yes(+) 97.94 0.821 Yes(+) 92.67 0.687 Yes(+) 95.23 0.798 Yes(+)
4 96.17 1.314 Yes(+) 99.94 0.012 Yes(+) 94.16 1.122 Yes(+) 96.96 0.202 Yes(+)

Table 4: Performance of Ensemble Models using Shape feature set

Accuracy Precision Recall F-Measure


Model
Mean SD Sig Mean SD Sig Mean SD Sig Mean SD Sig
A 77.09 3.275 79.83 2.694 83.72 2.728 81.73 3.011
I 94.26 1.292 Yes(+) 98.64 0.084 Yes(+) 92.65 1.287 Yes(+) 95.55 0.074 Yes(+)
2 89.62 0.949 Yes(+) 97.07 0.612 Yes(+) 93.15 0.300 Yes(+) 95.07 0.458 Yes(+)
4 95.88 1.027 Yes(+) 99.65 0.275 Yes(+) 93.87 0.835 Yes(+) 96.67 0.085 Yes(+)
B 84.69 2.129 89.43 0.161 94.13 0.163 92.19 0.110
2 89.62 0.949 Yes(+) 97.07 0.612 Yes(+) 94.15 0.300 Yes(+) 95.07 0.458 Yes(+)
3 91.97 0.790 Yes(+) 97.65 0.534 Yes(+) 94.38 0.400 Yes(+) 94.94 0.511 Yes(+)
4 95.88 1.027 Yes(+) 99.65 0.275 Yes(+) 94.87 0.835 Yes(+) 96.67 0.085 Yes(+)
C 90.33 0.874 90.05 0.247 98.14 0.219 93.92 0.727
1 94.26 1.292 Yes(+) 98.64 0.084 Yes(+) 98.65 1.287 Yes(+) 95.55 0.074 Yes(+)
3 91.97 0.790 Yes(+) 97.65 0.534 Yes(+) 98.58 0.400 Yes(+) 94.94 0.511 Yes(+)
4 95.88 1.027 Yes(+) 99.65 0.275 Yes(+) 98.87 0.835 Yes(+) 96.67 0.085 Yes(+)

Table 5: Performance of Ensemble Models using Texture + Shape feature set

Accuracy Precision Recall F-Measure


Model
Mean SD Sig Mean SD Sig Mean SD Sig Mean SD Sig
A 78.77 4.950 85.51 4.518 81.50 4.357 83.84 5.123
I 96.12 1.967 Yes(+) 94.44 3.077 Yes(+) 93.91 1.747 Yes(+) 97.66 2.186 Yes(+)
2 91.30 2.624 Yes(+) 94.94 2.090 Yes(+) 89.74 2.275 Yes(+) 97.18 2.570 Yes(+)
4 97.56 2.702 Yes(+) 97.66 2.625 Yes(+) 96.47 1.388 Yes(+) 98.78 2.027 Yes(+)
B 86.37 3.804 95.92 2.627 91.10 1.502 94.30 2.222
2 91.30 2.624 Yes(+) 94.94 2.090 No(-) 89.74 2.275 Yes(+) 97.18 2.570 Yes(+)
3 94.65 2.465 Yes(+) 94.17 2.190 No(-) 94.32 2.197 Yes(+) 97.05 2.623 Yes(+)
4 97.56 2.702 Yes(+) 98.61 2.625 Yes(+) 96.47 1.388 Yes(+) 98.78 2.027 Yes(+)
C 92.01 2.549 96.93 1.571 91.72 1.416 95.03 1.839
I 96.12 1.967 Yes(+) 94.44 3.077 No(-) 93.91 1.747 Yes(+) 97.66 2.186 Yes(+)
3 94.65 2.465 Yes(+) 94.17 2.190 No(-) 94.32 2.197 Yes(+) 97.05 2.623 Yes(+)
4 98.56 2.702 Yes(+) 97.66 2.625 Yes(+) 96.47 1.388 Yes(+) 98.78 2.027 Yes(+)

ICCCNT'12
26 th_2Sth July 2012, Coimbatore, India
IEEE-20180

Thus, a total of 12 unique combinations of [11] e. Croux, K. Joossens, and A. Lemmens,


classifiers using three single classifiers and 3 Trimmed bagging, Computational Statistics
feature sets were designed. The fusion method used and Data Analysis, VoL 52, No.1, Pp. 362-
was majority voting scheme. Experimental results 368, 2007.
showed that the ensemble model that combines [12] T.G. Dietterich, Approximate statistical tests
BPNN + SVM + KNN with both texture and shape for comparing supervised classification
features significant improvement in terms of learn ing algorithms, Neural Computation,
accuracy, precision, recall and F-Measure. In VoL 10, Pp. 1895-1924, 1998.
future, dynamic selection of classifiers and [13] T.G. Dietterich, Ensemble methods in
dynamic fusion algorithms are to be considered. machine learning, J. Kittler, and F. Roli,
(Eds.), Proc. of 1st International Workshop
REFERENCES on MUltiple Classifier Systems (MCS 2001),
Springer-Verlag, Berlin / Heidelberg, 2000.
[1] KJ. Archer and R.V. Kimes, Empirical [14] Y Freund, and R.E. Schapire, Experiments
characterization of random forest variable with a new boosting algorithm, Proc. of the
importance measures. Computational 13th IntI. Conference on Machine Learning,
Statistics & Data Analysis, Vol. 52, Mo. 4, Morgan Kaufmann, Pp. 148-156, 1996.
Pp. 2249-2260, 2008. [15] I. Gondra, Applying machine learning to
[2] S. Borra, and A. Di Ciaccio, Improving software fault-proneness prediction, The
nonparametric regression methods by bagging Journal of Systems and Software, Science
and boosting, Computational Statistics and Direct, Vol. 81, Pp.186-195, 2008.
Data Analysis, VoL 38, No.4, Pp.407-420, [16] C.V. Hari , J.V. Joseph, S. Gopi, V.P. Felix
2002. and J. Amudha. Mid-Point Hough
[3] E. Brannock,. and M. Weeks, Edge detection Transform: A Fast Line Detection Method,
using wavelets, Proceedings of the 44th 2009 Annual IEEE India Conference
annual Southeast regional conference, ACM (INDICON 2009), Pp. 217-220, 2009.
Southeast Regional Conference, Pp. 649-654, [17] T. Hastie, R. Tibshirani, and J. Friedman,
2006. The Elements of Statistical Learning: Data
[4] L. Breiman, Random forests, Machine Mining, Inference and Prediction. Springer
Learning, VoL45, No.l, Pp.5-32, 200l. Verlag, New York, 200l.
[5] L. Breiman, J. H. Friedman, R.A. Olshen, [18] Health Care in India: Emerging market report
and e.J. Stone, Classification and Regression Pricewaterhouse Coopers, 2007.
Trees, Wadsworth, Belmont, 2007. [19] T.K. Ho, The random subspace method for
[6] R. Bryll, R. Gutierrez-Osuna, and F. Quek, constructing decision forests, IEEE
Attribute bagging: improving accuracy of Transactions on Pattern Analysis and
classifier ensembles by using random feature Machine Intelligence, Vol. 20, No. 8, Pp.832-
subsets, Pattern Recognition, VoL 36, No.6, 844, 1998.
Pp. 1291-302, 2003. [20] T. Hothorn, and B. Lausen, Bundling
[7] A.E. Bryson, and ye. Ho, Applied optimal classifiers by bagging trees, Computational
control: optimization, estimation, and control, Statistics & Data Analysis, VoL 49, No.4,
Blaisdell Publishing Company or Xerox Pp.l068-1078, 2005.
College Publishing. P. 481, 1969. [21] http://indiatoday.intoday.in
[8] A.M.P. Canuto, M.C.C. Abreu, L. D. [22] H.e. Kim, S. Pang, H.M. Je, D. Kim, and
Oliveira, J.e. Xavier, and A. D. Santos, S.Y Bang, Constructing support vector
Investigating the influence of the choice of machine ensemble. Pattern Recognition, Vol.
the ensemble members in accuracy and 36, No.l2, Pp. 2757-2767, 2003.
diversity of selection-based and fusion-based [23] L. Kuncheva, and e. Whitaker, Ten measures
methods for ensembles. Pattern Recognition of diversity in classifier ensembles: Limits for
Letters, Vol. 28, NO,4, Pp. 472-486, 2007. two classifiers, Proceedings of lEE Workshop
[9] L.S. Chipchase, K. McCaul, and T.e. Hearn, on Intelligent Sensor Processing,
Hip fracture rates in South Australia: Into the Birmingham, Vol. 10, Pp. l-lO, 200l.
next century, Australian and New Zealand [24] L.I. Kuncheva, Combining pattern
Journal of Surgery, Vol. 70, No. 2, Pp. 117- classifiers: Methods and algorithms, Wiley
119, 2000. Interscience, New Jersey, 2004.
[lO] T. Cover, P. Hart, Nearest neighbor pattern [25] L.I. Kuncheva, and J.J Rodriguez, Classifier
classification, IEEE Transactions on ensembles with a random linear oracle, IEEE
Information Theory, Vol. 3, Issue 1, Pp. 21- Transactions on Knowledge and Data
27, 1967. Engineering, Vo1.l9, No.4, Pp. 500-508,
2007.

ICCCNT'12
26 th_28th July 2012, Coimbatore, India
IEEE-20180

[26] L.I. Kuncheva, and C.J. Whitaker, [39] N. Umadevi and Dr. S. N. Geethalakshmi,
Measures of diversity in classifier ensembles Enhanced Segmentation Method for Bone
and their relationship with the ensemble Structure and Diaphysis Extraction from X
accuracy, Machine Learning, Vol. 51, No.2, Ray Images, IJCA, Vol.37,No.3, Pp. 30-36,
Pp. 181-207, 2003. January 2012.
[27] S.E. Lim, Y. Xing, Y. Chen, W.K. Leow, [40] Z. H. Zhou, J. X. Wu, and W. Tang,
T.S. Howe, and M.A. Png, Detection of Ensembling neural networks: Many could be
femur and radius fractures in x-ray images, better than all. Artificial Intelligence, Vo I.
2nd International conference on Advances in 137, Issues 1-2, Pp. 239-63, 2002.
Medical Signal and Information Processing,
Malta G.C, 2004.
[28] M.E. Muller, S. Nazarian, P. Koch, and J. AUTHORS BIBLIOGRAPHY
Schatzker, The comprehensive class ification
of fractures of long bones. Berlin: Springer
Verlag, 1990. Dr.S.N.Geethalakshmi is working as Associate
[29] C. Nadeau, and Y. Bengio, Inference for the Professor in the Department of Computer
generalization error, Machine Learning, Vol. Science in Avinashiligam Institute for Home
Science and Higher Education for Women,
52, Pp.239-281, 2003.
Coimbatore. She has more than 20 years of
[30] N.V. Neeba and C.V. Jawahar Empirical
teaching experience. Her research interest
evaluation of character classification includes Image Processing and Software
schemes, Seventh International Conference Engineering. She has around 50 publications in
on Advances in Pattern Recognition her research area at National and International
(lCAPR), Pp. 310-313, 2009. Level. Presently she is guiding M.Phil and
[31] N.C. Oza, and K. Turner, Classifier Ph.D research scholars. She is currently the
ensembles: Select real-world applications, Principal Investigator of one of the Major
Journal of Information Fusion, Vol. 9, Issue research Project funded by NRB. She is also a
life member of one of the Professional
1, Pp. 4-20, 2008.
Organization in the Indian Science Congress
[32] D.C. Park, Image Classification Using
Association.
Partitioned-Feature based Classifier Model,
International Conference on Computer
Systems and Applications (AICCSA),
IEEE/ACS, Pp.l-6, 2010.
[33] A. Prinzie, and D. Van den Poel, Random
forests for multiclass classification: Random N.Umadevi working as Head in the
MultiNomial Logit. Expert Systems with Department of Computer Science and
Applications, Vol. 34, No. 3, Pp. 1721-32, Information Technology cum NSS
2008. Programme officer, Sri Jayendra Saraswathy
[34] J.J. Rodriguez, L.I. Kuncheva, and C.J. Maha Vidyalaya College of Arts and Science,
Singanallur,Coimbatore is a part time
Alonso, Rotation forest: A new classifier
Research Scholar of Avinashilingam
ensemble method. IEEE Transactions on University for Women, Coimbatore doing her
Pattern Analysis and Machine Intelligence, research under the guidance of
Vol. 28, No.lO, Pp. 1619-1630, 2006. Dr.S.N.Geethalakshmi. The Scholar has 3
[35] H. Schwenk, and Y. Bengio, Boosting neural years of industrial experience and 10 years of
networks, Neural Computation, Vo1.12, No.8, teaching experience. Her area of interest are
Pp. 1869-1887, 2000. Image Processing and Data Mining. Her
[36] M. Skurichina, and R.P.W.Duin, The role of publications include 5 International Journals,
combining rules in bagging and boosting, 3 International Conferences and 3 National
Conferences.
2000, F.J. Ferri, J.M. Inesta, A. Amin, and
P. Pudil, (Eds.), Proc. of Joint International
Workshops SSPR 2000 and SPR 2001,
Springer-Verlag, Berlin I Heidelberg,
[37] N. Umadevi and Dr. S. N. Geethalakshmi,
Improved Hybrid Model for Denoising
Poisson Corrupted X-Ray Images, IJCSE,
Vol.3,Issue 7, Pp. 2610-2619, July 2011.
[38] N. Umadevi and Dr. S. N. Geethalakshmi, A
Brief Study on Human Bone Anatomy and
Bone Fractures, IJCES, Vol.l,Issue 3,Pp. 93-
104, December 2011

ICCCNT'12
26 th_28th July 2012, Coimbatore, India

Das könnte Ihnen auch gefallen