Titlebook: Intelligent Data Engineering and Automated Learning -- IDEAL 2014; 15th International C Emilio Corchado,José A. Lozano,Hujun Yin Conference

hazard

Diversified Random Forests Using Random Subspaces,, giving a weight to each subspace according to its predictive power, and using this weight in majority voting. Experimental study on 15 real datasets showed favourable results, demonstrating the potential of the proposed method.

同義聯(lián)想法

Multi-step Forecast Based on Modified Neural Gas Mixture Autoregressive Model,g patterns and its suitability for various step-ahead predictions. Experimental results on several financial time series and benchmark data demonstrate the effectiveness of proposed method and markedly improvement performances over many existing neural networks.

indubitable

LBP and Machine Learning for Diabetic Retinopathy Detection, patterns (LBP) to extract local features, while in the second stage, we have applied artificial neural networks, random forest and support vector machines for the detection task. Preliminary results show that random forest was the best classifier with 97.46% of accuracy, using a data set of 71 images.

Carcinogen

Jubilation

柏樹

entreat

Object-Neighbourhood Clustering Ensemble Method,tasets. The results show that our ensemble method outperforms the co-association method, when the Average linkage is used. Furthermore, the results show that our ensemble method is more accurate than the baseline algorithm, and this indicates that the clustering ensemble method is more consistent and reliable than a single clustering algorithm.

摻假

TOXIN

Alpha-Cells

0302-9743 optimization, regression, classification, clustering, biological data processing, text processing, and image/video analysis.978-3-319-10839-1978-3-319-10840-7Series ISSN 0302-9743 Series E-ISSN 1611-3349