Results 1  10
of
18
M.: A simple approach to ordinal classification
 In: Proc 12th Europ Conf on Machine Learning
, 2001
"... Abstract. Machine learning methods for classification problems commonly assume that the class values are unordered. However, in many practical applications the class values do exhibit a natural order—for example, when learning how to grade. The standard approach to ordinal classification converts th ..."
Abstract

Cited by 85 (3 self)
 Add to MetaCart
(Show Context)
Abstract. Machine learning methods for classification problems commonly assume that the class values are unordered. However, in many practical applications the class values do exhibit a natural order—for example, when learning how to grade. The standard approach to ordinal classification converts the class value into a numeric quantity and applies a regression learner to the transformed data, translating the output back into a discrete class value in a postprocessing step. A disadvantage of this method is that it can only be applied in conjunction with a regression scheme. In this paper we present a simple method that enables standard classification algorithms to make use of ordering information in class attributes. By applying it in conjunction with a decision tree learner we show that it outperforms the naive approach, which treats the class values as an unordered set. Compared to specialpurpose algorithms for ordinal classification our method has the advantage that it can be applied without any modification to the underlying learning scheme. 1
M (2003) Using evolutionary algorithms as instance selection for data reduction in KDD: an experimental study
 IEEE Trans Evol Comput
"... Abstract—Evolutionary algorithms are adaptive methods based on natural evolution that may be used for search and optimization. As data reduction in knowledge discovery in databases (KDDs) can be viewed as a search problem, it could be solved using evolutionary algorithms (EAs). In this paper, we hav ..."
Abstract

Cited by 68 (20 self)
 Add to MetaCart
(Show Context)
Abstract—Evolutionary algorithms are adaptive methods based on natural evolution that may be used for search and optimization. As data reduction in knowledge discovery in databases (KDDs) can be viewed as a search problem, it could be solved using evolutionary algorithms (EAs). In this paper, we have carried out an empirical study of the performance of four representative EA models in which we have taken into account two different instance selection perspectives, the prototype selection and the training set selection for data reduction in KDD. This paper includes a comparison between these algorithms and other nonevolutionary instance selection algorithms. The results show that the evolutionary instance selection algorithms consistently outperform the nonevolutionary ones, the main advantages being: better instance reduction rates, higher classification accuracy, and models that are easier to interpret. Index Terms—Data mining (DM), data reduction, evolutionary algorithms (EAs), instance selection, knowledge discovery.
A twolevel learning method for generalized multiinstance problems
 In Proceedings of the Fourteenth European Conference on Machine Learning
, 2003
"... Abstract. In traditional multiinstance (MI) learning, a single positive instance in a bag produces a positive class label. Hence, the learner knows how the bag’s class label depends on the labels of the instances in the bag and can explicitly use this information to solve the learning task. In this ..."
Abstract

Cited by 48 (5 self)
 Add to MetaCart
(Show Context)
Abstract. In traditional multiinstance (MI) learning, a single positive instance in a bag produces a positive class label. Hence, the learner knows how the bag’s class label depends on the labels of the instances in the bag and can explicitly use this information to solve the learning task. In this paper we investigate a generalized view of the MI problem where this simple assumption no longer holds. We assume that an “interaction” between instances in a bag determines the class label. Our twolevel learning method for this type of problem transforms an MI bag into a single metainstance that can be learned by a standard propositional method. The metainstance indicates which regions in the instance space are covered by instances of the bag. Results on both artificial and realworld data show that this twolevel classification approach is well suited for generalized MI problems. 1
Machine learning methods for predicting failures in hard drives: A multiple instance application
, 2005
"... We compare machine learning methods applied to a difficult realworld problem: predicting computer harddrive failure using attributes monitored internally by individual drives. The problem is one of detecting rare events in a time series of noisy and nonparametricallydistributed data. We develop ..."
Abstract

Cited by 39 (1 self)
 Add to MetaCart
(Show Context)
We compare machine learning methods applied to a difficult realworld problem: predicting computer harddrive failure using attributes monitored internally by individual drives. The problem is one of detecting rare events in a time series of noisy and nonparametricallydistributed data. We develop a new algorithm based on the multipleinstance learning framework and the naive Bayesian classifier (miNB) which is specifically designed for the low falsealarm case, and is shown to have promising performance. Other methods compared are support vector machines (SVMs), unsupervised clustering, and nonparametric statistical tests (ranksum and reverse arrangements). The failureprediction performance of the SVM, ranksum and miNB algorithm is considerably better than the threshold method currently implemented in drives, while maintaining low false alarm rates. Our results suggest that nonparametric statistical tests should be considered for learning problems involving detecting rare events in time series data. An appendix details the calculation of ranksum significance probabilities in the case of discrete, tied observations, and we give new recommendations about when the exact calculation should be used instead of the commonlyused normal approximation. These normal approximations may be particularly inaccurate for rare event problems like hard drive failures.
On the combination of evolutionary algorithms and stratified strategies for training set selection in data mining
 Applied Soft Computing
, 2006
"... Evolutionary algorithms are adaptive methods based on natural evolution that may be used for search and optimization. As Training Set Selection can be viewed as a search problem, it could be solved using evolutionary algorithms. In this paper, we have carried out an empirical study of the performanc ..."
Abstract

Cited by 19 (0 self)
 Add to MetaCart
(Show Context)
Evolutionary algorithms are adaptive methods based on natural evolution that may be used for search and optimization. As Training Set Selection can be viewed as a search problem, it could be solved using evolutionary algorithms. In this paper, we have carried out an empirical study of the performance of CHC as representative evolutionary algorithm model. This study includes a comparison between this algorithm and other nonevolutionary instance selection algorithms applied in large size data sets for Training Set Selection. The results show that the stratified evolutionary instance selection algorithms consistently outperform the nonevolutionary ones, the main advantages being: better instance reduction rates, higher classification accuracy and models that are easier to interpret. 1
Applying propositional learning algorithms to multiinstance data
, 2003
"... data ..."
(Show Context)
Converting numerical classification into text classification
 ARTIFICIAL INTELLIGENCE
, 2003
"... Consider a supervised learning problem in which examples contain both numerical and textvalued features. To use traditional featurevectorbased learning methods, one could treat the presence or absence of a word as a Boolean feature and use these binaryvalued features together with the numerical ..."
Abstract

Cited by 7 (2 self)
 Add to MetaCart
Consider a supervised learning problem in which examples contain both numerical and textvalued features. To use traditional featurevectorbased learning methods, one could treat the presence or absence of a word as a Boolean feature and use these binaryvalued features together with the numerical features. However, the use of a textclassification system on this is a bit more problematicin the most straightforward approach each number would be considered a distinct token and treated as a word. This paper presents an alternative approach for the use of text classification methods for supervised learning problems with numericalvalued features in which the numerical features are converted into bagofwords features, thereby making them directly usable by text classification methods. We show that even on purely numericalvalued data the results of text classification on the derived textlike representation outperforms the more naive numbersastokens representation and, more importantly, is competitive with mature numerical classification methods such as C4.5, Ripper, and SVM. We further show that on mixedmode data adding numerical features using our approach can improve performance over not adding those features.
Unsupervised Discretization using Treebased Density Estimation
"... Abstract. This paper presents an unsupervised discretization method that performs density estimation for univariate data. The subintervals that the discretization produces can be used as the bins of a histogram. Histograms are a very simple and broadly understood means for displaying data, and our m ..."
Abstract

Cited by 4 (0 self)
 Add to MetaCart
(Show Context)
Abstract. This paper presents an unsupervised discretization method that performs density estimation for univariate data. The subintervals that the discretization produces can be used as the bins of a histogram. Histograms are a very simple and broadly understood means for displaying data, and our method automatically adapts bin widths to the data. It uses the loglikelihood as the scoring function to select cut points and the crossvalidated loglikelihood to select the number of intervals. We compare this method with equalwidth discretization where we also select the number of bins using the crossvalidated loglikelihood and with equalfrequency discretization. 1
Using Text Classifiers for Numerical Classification
 In Proceedings of the Seventeenth International Joint Conference on Artificial Intelligence, Seattle,WA,August
, 2001
"... Consider a supervised learning problem in which examples contain both numerical and textvalued features. To use traditional featurevector based learning methods, one could treat the presence or absence of a word as a Boolean feature and use these binaryvalued features together with the numerical ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
Consider a supervised learning problem in which examples contain both numerical and textvalued features. To use traditional featurevector based learning methods, one could treat the presence or absence of a word as a Boolean feature and use these binaryvalued features together with the numerical features. However, the use of a textclassification system on this is a bit more problematic  in the most straightforward approach each number would be considered a distinct token and treated as a word. This paper presents an alternative approach for the use of text classification methods for supervised learning problems with numericalvalued features in which the numerical features are converted into bagofwords features, thereby making them directly usable by text classification methods. We show that even on purely numericalvalued data the results of textclassification on the derived textlike representation outperforms the more naive numbersastokens representation and, more importantly, is competitive with mature numerical classification methods such as C4.5 and Ripper. 1