DocumentCode :
2702967
Title :
Selecting diverse members of neural network ensembles
Author :
Navone, H.D. ; Verdes, P.F. ; Granitto, P.M. ; Ceccatto, H.A.
Author_Institution :
Inst. de Fisica Rosario, Rosario, Argentina
fYear :
2000
fDate :
2000
Firstpage :
255
Lastpage :
260
Abstract :
Ensembles of artificial neural networks have been used as classification/regression machines, showing improved generalization capabilities that outperform those of single networks. However, it has been recognized that for aggregation to be effective the individual network must be as accurate and diverse as possible. An important problem is, then, how to choose the aggregate members in order to have an optimal compromise between these two conflicting conditions. We propose here a new method for selecting members of regression/classification ensembles that leads to small aggregates with few but very diverse individual predictors. Using artificial neural networks as individual learners, the algorithm is favorably tested against other methods in the literature, producing a remarkable performance improvement on the standard statistical databases used as benchmarks. In addition, and as a concrete application, we study the sunspot time series and predict the remaining of the current cycle 23 of solar activity
Keywords :
astronomy computing; generalisation (artificial intelligence); learning (artificial intelligence); neural nets; pattern classification; sunspots; time series; aggregation; generalization; learning; neural networks; pattern classification; regression machines; solar activity; sunspot; time series; Aggregates; Artificial neural networks; Bagging; Benchmark testing; Concrete; Databases; Diversity reception; Neural networks; Noise generators; Optimization methods;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks, 2000. Proceedings. Sixth Brazilian Symposium on
Conference_Location :
Rio de Janeiro, RJ
ISSN :
1522-4899
Print_ISBN :
0-7695-0856-1
Type :
conf
DOI :
10.1109/SBRN.2000.889748
Filename :
889748
Link To Document :
بازگشت