چكيده به لاتين
There are many classification algorithms in data mining. On the other hand, it can’t be said that a special algorithm is the best, since the performance and accuracy of different algorithms differ from one set to another according to their characteristics. So classification algorithm selection is one of the most important challenges in data mining. By combining set of the algorithms, it is possible to improve the performance and accuracy of the result of the classification, as compared to each algorithm.
In this study we proposed a method based on meta-learning and Ensemble methods. This method aims to reduce the selection time by automatic recommendation of the best classifier combination for a given dataset considering it’s meta-features.
For evaluation, we compared the error rate of proposed method with the average error rate of individual classifiers for several unseen datasets. One of these unseen datasets is named Refractive error. Uncorrected refractive errors are the first reason of low-vision and the second cause of curable blindness in the world. Data mining can, therefore, be used as an effective method to determine the causes of refractive errors.
In this study, we will introduce the risk factors of refractive errors, by applying the model presented on the refractive errors dataset.