C4.5: programs for machine learning
C4.5: programs for machine learning
Artificial Intelligence Review - Special issue on lazy learning
Control-Sensitive Feature Selection for Lazy Learners
Artificial Intelligence Review - Special issue on lazy learning
Advances in knowledge discovery and data mining
Advances in knowledge discovery and data mining
Improving Minority Class Prediction Using Case-Specific Feature Weights
ICML '97 Proceedings of the Fourteenth International Conference on Machine Learning
A Dynamic Integration Algorithm for an Ensemble of Classifiers
ISMIS '99 Proceedings of the 11th International Symposium on Foundations of Intelligent Systems
Decomposition of Heterogeneous Classification Problems
IDA '97 Proceedings of the Second International Symposium on Advances in Intelligent Data Analysis, Reasoning about Data
Learning Feature Selection for Medical Databases
CBMS '99 Proceedings of the 12th IEEE Symposium on Computer-Based Medical Systems
Data Mining using MLC++, A Machine Learning Library in C++
ICTAI '96 Proceedings of the 8th International Conference on Tools with Artificial Intelligence
A study of cross-validation and bootstrap for accuracy estimation and model selection
IJCAI'95 Proceedings of the 14th international joint conference on Artificial intelligence - Volume 2
| Hi-index | 0.02 |
Multidimensional data is often feature-space heterogeneous so that different features have different importance in different subareas of the whole space. In this paper we suggest a technique that searches for a strategic splitting of the feature space identifying the best subsets of features for each instance. Our technique is based on the wrapper approach where a classification algorithm is used as the evaluation function to differentiate between several feature subsets. In order to make the feature selection local, we apply the recently developed technique for dynamic integration of classifiers. It allows us to determine what classifier and with what feature subset should be applied for each new instance. In order to restrict the number of feature combinations being analyzed we propose to use decision trees. For each test instance we consider only those feature combinations that include features present in the path taken by the test instance in the decision tree built on the whole feature set. We evaluate our technique on datasets from the UCI machine learning repository. In our experiments, we use the C4.5 algorithm as the learning algorithm for base classifiers and for decision trees that guide the local feature selection. The experiments show advantages of the local feature selection in comparison with the selection of one feature subset for the whole space.