Classification of imbalanced datasets using naive bayes
Imbalanced data set had tendency to effect classifier performance in machine learning due to the greater influence given by majority data that overlooked the minority ones. But in classifying data, more important class is given by the minority data. In order to solve this problem, original Naïve Bay...
保存先:
| 第一著者: | |
|---|---|
| フォーマット: | 学位論文 |
| 言語: | English |
| 出版事項: |
2011
|
| 主題: | |
| オンライン･アクセス: | http://eprints.utm.my/id/eprint/31941/5/NurMaisarahMohdSobranMFKE2011.pdf http://eprints.utm.my/id/eprint/31941/ |
| タグ: |
タグ追加
タグなし, このレコードへの初めてのタグを付けませんか!
|
| 要約: | Imbalanced data set had tendency to effect classifier performance in machine learning due to the greater influence given by majority data that overlooked the minority ones. But in classifying data, more important class is given by the minority data. In order to solve this problem, original Naïve Bayes was purposed as classifier for imbalanced data set. Our main interest is to investigate the performance of original Naïve Bayes classifier in imbalanced datasets. From the four UCI imbalanced datasets that been used, the purposed techniques show that, Naïve Bayes doing well in Herbaman’s datasets and satisfying results in other datasets. |
|---|