KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 7 of 7

Some of the metrics are blocked by your
consent settings
Min-uncertainty & max-certainty criteria of neighborhood rough-mutual feature selection
(2017-01-01)
Foithong, Sombut
;
Srinil, Phaitoon
;
Pinngern, Ouen
;
Attachoo, Boonwat
Feature Selection (FS) is viewed as an important preprocessing step for pattern recognition, machine learning, and data mining. Most existing FS methods based on rough set theory use the dependency function for evaluating the goodness of a feature subset. However, these FS methods may unsuccessfully be applied on dataset with noise, which determine only information from a positive region but neglect a boundary region. This paper proposes a criterion of the maximal lower approximation information (Max-Certainty) and minimal boundary region information (Min-Uncertainty), based on neighborhood rough set and mutual information for evaluating the goodness of a feature subset. We combine this proposed criterion with neighborhood rough set, which is directly applicable to numerical and heterogeneous features, without involving a discretization of numerical features. Comparing it with the rough set based approaches, our proposed method improves accuracy over various experimental data sets. Experimental results illustrate that much valuable information can be extracted by using this idea. This proposed technique is demonstrated on discrete, continuous, and heterogeneous data, and is compared with other FS methods in terms of subset size and classification accuracy.
Some of the metrics are blocked by your
consent settings
Rough-mutual feature selection based on min-uncertainty and max-certainty
(2012-01-01)
Foitong, Sombut
;
Pinngern, Ouen
;
Attachoo, Boonwat
Feature selection (FS) plays an important role in pattern recognition and machine learning. FS is applied to dimensionality reduction and its purpose is to select a subset of the original features of a data set which is rich in the most useful information. Most existing FS methods based on rough set theory focus on dependency function, which is based on lower approximation as for evaluating the goodness of a feature subset. However, by determining only information from a positive region but neglecting a boundary region, most relevant information could be invisible. This paper, the maximal lower approximation (Max Certainty) minimal boundary region (Mm Uncertainty) criterion, focuses on feature selection methods based on rough set and mutual infonnation which use different values among the lower approximation information and the information contained in the boundary region. The use of this idea can result in higher predictive accuracy than those obtained using the measure based on the positive region (certainty region) alone. This demonstrates that much valuable information can be extracted by using this idea. Experimental results are illustrated for discrete, continuous, and microarray data and compared with other FS methods in terms of subset size and classification accuracy. key words: rough sets, mutual information, feature selection, boundary region, classification accuracy. Copyright © 2012 The Institute of Electronics, Information and Communication Engineers.
Some of the metrics are blocked by your
consent settings
Cost-sensitive XCS classifier system addressing imbalance problems
(2008-12-01)
Thach, Nguyen Huy
;
Rojanavasu, Porntep
;
Pinngern, Ouen
The class imbalance problem has been recognized as a crucial problem in machine learning and data mining. Learning systems tend to be biased towards the majority class and thus have poor generalization for the minority class instances. This paper analyses the imbalance problem in accuracy-based learning classifier systems. In particular, we propose a novel approach based on XCS classifier system and cost-sensitive learning. In our approach, the reward value of correctly identifying the positive (rare) class outweighs the value of correctly identifying the common class. This research provides guidelines to set reward base on the dataset imbalance ratio and a method to calculate reward online base on the information collected by XCS during training is also proposed. Experimental results on synthetic and real-life datasets show that, with appropriate reward settings, XCS is robust to class imbalances. © 2008 IEEE.
Some of the metrics are blocked by your
consent settings
Towards adapting XCS for imbalance problems
(2008-12-01)
Nguyen, Thach Huy
;
Foitong, Sombut
;
Srinil, Phaitoon
;
Pinngern, Ouen
The class imbalance problem has been recognized as a crucial problem in machine learning and data mining. Learning systems tend to be biased towards the majority class and thus have poor performance in classifying the minority class instances. This paper analyzes the imbalance problem in accuracy-based learning classifier system XCS. XCS has shown excellent performance on some data mining tasks, but as other classifiers, it also performs poorly on imbalance data problems. We analyze XCS's behavior on various imbalance levels and propose an appropriate parameter tuning to improve performance of the system. Particularly, XCS is adapted to eliminate over-general classifiers and protect accurate classifiers of minority class. Experimental results in Boolean function problems show that, with proposal adaptations, XCS is robust to class imbalance. © 2008 Springer Berlin Heidelberg.
Some of the metrics are blocked by your
consent settings
Rough set and XCS in classification problems
(2008-09-22)
Nguyen, Thach H.
;
Foitong, Sombut
;
Pinngern, Ouen
XCS is known to degrade in classification performance when faced with many features that are redundant for rules discovery. In this paper, we propose a novel system combining of rough sets and XCS to deal with the mentioned problem. Firstly, rough set theory is used to handle inconsistent input datasets. The purpose of feature reduction by rough set is to identify the most significant attributes and eliminate the irrelevant ones to form a good feature subset for classification. Secondly, the reduced datasets are used to create a set of rules by using XCS. The main contribution of XCS to learning theory is its rules generation without experts. Finally, by applying the set of rules, we can classify unseen datasets into their specific classes. Experimental results on real-life datasets show that the proposed method can reduce storage space as well as can preserve and may also improve solution accuracy. Beside that, the rule retrieval time is also greatly reduced because the use of Rough-XCS classifier contains a smaller amount of instances with fewer features. Furthermore, the proposed method has a high potential to be used as a mean to construct a classifier system that copes with incomplete, noisy and chaotic data. ©2008 IEEE.
Some of the metrics are blocked by your
consent settings
Spartan simplicity: A pruning algorithm for neural nets
(2008-08-01)
Jearanaitanakij, Kietikul
;
Pinngern, Ouen
Having more hidden units than necessary can produce a neural network that has a poor generalization. This paper proposes a new algorithm for pruning unnecessary hidden units away from the single-hidden layer feedforward neural networks, resulting in a Spartan network. Our approach is simple and easy to implement, yet produces a very good result. The idea is to train the network until it begins to lose its generalization. Then the algorithm measures the sensitivity and automatically prunes away the most irrelevant unit. We define this sensitivity as the absolute difference between the desirable output and the output of the pruned network. Unlike other pruning methods, our algorithm is distinct in calculating the sensitivity from the validation set, instead of the training set, without increasing the asymptotic time complexity of the back-propagation algorithm. In addition, for a classification problem, we raise a point that the sensitivities of some well-known pruning algorithms may still underestimate the irrelevance of hidden unit even though the validation set is used in measuring the sensitivity. We resolve this problem by considering the number of misclassified patterns as the main concern. The Spartan simplicity algorithm is applied to three artificial and seven standard benchmarks. In most problems, the algorithm can produce a compact-sized network with high generalization ability in comparison with other pruning algorithms. © 2008 World Scientific Publishing Company.
Some of the metrics are blocked by your
consent settings
An information gain technique for acceleration of convergence of artificial neural networks
(2005-12-01)
Jearanaitanakij, Kietikul
;
Pinngern, Ouen
This paper presents an application of information gain to accelerate the convergence time of Artificial Neural Networks (ANNs). We improve Hagiwara's convergence acceleration algorithm by applying information gain to it. The first step of our proposed technique is to calculate information gains of all features (or attributes) in training data and pass those gains through all hidden units in the next layer. During the training process, the algorithm monitors sum-squared error at the output layer. When the variation of sum-squared error becomes small, the worst hidden unit is detected. Next, all the weights connected to the worst hidden unit are reset to random values within the appropriate ranges. These ranges are determined by the propagated information gain of the worst hidden unit. Then, the network is retrained. When the number of weight resetting trials reaches a certain number, a new hidden unit is added to the network and the whole training process is repeated. Our experimental results on standard benchmarks show remarkable outputs in terms of convergence time. © 2005 IEEE.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results