Title :
Active Meta-Learning with Uncertainty Sampling and Outlier Detection
Author :
Prudêncio, Ricardo B C ; Ludermir, Teresa B.
Author_Institution :
Dept. of Inf. Sci., Fed. Univ. of Pernambuco, Recife
Abstract :
Meta-Learning has been used to predict the performance of learning algorithms based on descriptive features of the learning problems. Each training example in this context, i.e. each meta-example, stores the features of a given problem and information about the empirical performance obtained by the candidate algorithms on that problem. The process of constructing a set of meta-examples may be expensive, since for each problem avaliable for meta-example generation, it is necessary to perform an empirical evaluation of the candidate algorithms. Active Meta-Learning has been proposed to overcome this limitation by selecting only the most informative problems in the meta-example generation. In this work, we proposed an Active Meta-Learning method which combines Uncertainty Sampling and Outlier Detection techniques. Experiments were performed in a case study, yielding significant improvement in the Meta-Learning performance.
Keywords :
learning (artificial intelligence); sampling methods; active meta-learning algorithm; outlier detection technique; uncertainty sampling technique; Costs; Helium; Information science; Learning systems; Machine learning; Machine learning algorithms; Performance evaluation; Prediction algorithms; Sampling methods; Uncertainty;
Conference_Titel :
Neural Networks, 2008. IJCNN 2008. (IEEE World Congress on Computational Intelligence). IEEE International Joint Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-1820-6
Electronic_ISBN :
1098-7576
DOI :
10.1109/IJCNN.2008.4633815