![Page 1: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/1.jpg)
RUSBoost : A Hybrid Approach to Alleviating Class Imbalance
19. 05. 17Yongwon Jo
Data Mining & Quality Analytics Lab.
![Page 2: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/2.jpg)
Contents
I. Introduction to Class Imbalance problem
II. How to solve Class Imbalance problem
III. RUSBoost vs. SMOTEBoost
IV.Result of experiments
V. Conclusion
![Page 3: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/3.jpg)
Contents
I. Introduction to Class Imbalance problem
II. How to solve Class Imbalance problem
III. RUSBoost vs. SMOTEBoost
IV.Result of experiments
V. Conclusion
![Page 4: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/4.jpg)
I. Introduction to Class Imbalance problem
Class Imbalance problem
It is the problem in classification where the total number of a class of data (positive) is far less than the total number of another class of data (negative).
This problem exists for many domains.
Heart Disease
Health
Heart Disease
Normal
Spam
![Page 5: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/5.jpg)
I. Introduction to Class Imbalance problem
Class Imbalance problem
Below plots are the class imbalance situation I actually saw.
![Page 6: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/6.jpg)
I. Introduction to Class Imbalance problem
Class Imbalance problem
Below plots are the class imbalance situation I actually saw.
It is a bar that shows the output quantity divided by the remain quantity.
![Page 7: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/7.jpg)
I. Introduction to Class Imbalance problem
Class Imbalance problem
Below plots are the class imbalance situation I actually saw.
It is a bar chart about whether the lot will be put into the next process.
![Page 8: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/8.jpg)
I. Introduction to Class Imbalance problem
Class Imbalance problem
RandomForestClassifier(max_depth=30, n_estimators=200)
Train dataset -> Accuracy : 0.90089 | F1 : 0.652276 | Recall : 0.98723 | Precision : 0.45484
Validation dataset -> Accuracy : 0.83854 | F1 : 0.19458 | Recall : 0.29088 | Precision : 0.14618
![Page 9: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/9.jpg)
I. Introduction to Class Imbalance problem
Class Imbalance problem
Predicted
0 1 2 3 4 5 6 7 8 9 10
Actual
0 515126 86 1 6 8 7 3 2 4 2 8572
1 6900 528 17 2 0 0 2 0 0 2 1585
2 5375 147 20 8 3 0 0 0 1 0 1596
3 4154 50 11 9 1 1 2 0 1 0 1535
4 3808 37 5 5 3 1 0 0 0 0 1436
5 3407 27 1 2 1 0 1 0 1 1 1450
6 2870 16 0 1 0 2 2 1 0 0 1370
7 2928 11 1 0 0 0 1 1 0 0 1464
8 2596 18 0 3 2 0 0 0 1 1 1383
9 2742 7 0 0 0 2 1 0 1 0 1466
10 56073 37 6 0 4 3 4 4 4 4 71017
RandomForestClassifier(max_depth=30, n_estimators=200)
Train dataset -> Accuracy : 0.90089 | F1 : 0.652276 | Recall : 0.98723 | Precision : 0.45484
Validation dataset -> Accuracy : 0.83854 | F1 : 0.19458 | Recall : 0.29088 | Precision : 0.14618
![Page 10: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/10.jpg)
RandomForestClassifier(max_depth=30, n_estimators=200)
Train dataset -> Accuracy : 0.90089 | F1 : 0.652276 | Recall : 0.98723 | Precision : 0.45484
Validation dataset -> Accuracy : 0.83854 | F1 : 0.19458 | Recall : 0.29088 | Precision : 0.14618
I. Introduction to Class Imbalance problem
Class Imbalance problem
Predicted
0 1 2 3 4 5 6 7 8 9 10
Actual
0 515126 86 1 6 8 7 3 2 4 2 8572
1 6900 528 17 2 0 0 2 0 0 2 1585
2 5375 147 20 8 3 0 0 0 1 0 1596
3 4154 50 11 9 1 1 2 0 1 0 1535
4 3808 37 5 5 3 1 0 0 0 0 1436
5 3407 27 1 2 1 0 1 0 1 1 1450
6 2870 16 0 1 0 2 2 1 0 0 1370
7 2928 11 1 0 0 0 1 1 0 0 1464
8 2596 18 0 3 2 0 0 0 1 1 1383
9 2742 7 0 0 0 2 1 0 1 0 1466
10 56073 37 6 0 4 3 4 4 4 4 71017
![Page 11: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/11.jpg)
Contents
I. Introduction to Class Imbalance problem
II. How to solve Class Imbalance problem
III. RUSBoost vs. SMOTEBoost
IV.Result of experiments
V. Conclusion
![Page 12: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/12.jpg)
II. How to solve Class Imbalance problem
Sampling Techniques
Algorithm Techniques
Feature selection Techniques
Over Sampling vs. Under Sampling
AdaBoost ,……
Lots of feature selection techniques
출처 : Class Imbalance Problem in Data Mining: Review 1Mr.Rushi Longadge, 2 Ms. Snehlata S. Dongre, 3Dr. Latesh Malik
![Page 13: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/13.jpg)
II. How to solve Class Imbalance problem
Sampling Techniques
Algorithm Techniques
Feature selection Techniques
Over Sampling vs. Under Sampling
Boosting, AdaBoost,
Lots of feature selection techniques
![Page 14: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/14.jpg)
II. How to solve Class Imbalance problem
Over Sampling
① How to create repeatedly instances of positive class.
② SMOTE : Synthetic Minority Over-Sampling
Positive
Negative
Feature Space
![Page 15: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/15.jpg)
II. How to solve Class Imbalance problem
Over Sampling
① How to create repeatedly instances of positive class.
② SMOTE : Synthetic Minority Over-Sampling
Positive
Negative
Feature Space
![Page 16: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/16.jpg)
II. How to solve Class Imbalance problem
Over Sampling
① How to create repeatedly instances of positive class.
② SMOTE : Synthetic Minority Over-Sampling
Positive
Negative
Feature Space
Pred
![Page 17: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/17.jpg)
II. How to solve Class Imbalance problem
Over Sampling
① How to create repeatedly instances of positive class.
② SMOTE : Synthetic Minority Over-Sampling
Positive
Negative
Feature Space
True Pred
![Page 18: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/18.jpg)
II. How to solve Class Imbalance problem
Over Sampling
① How to create repeatedly instances of positive class.
② SMOTE : Synthetic Minority Over-Sampling
Positive
Negative
Feature Space
True Pred
![Page 19: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/19.jpg)
II. How to solve Class Imbalance problem
Over Sampling
① How to create repeatedly instances of positive class.
② SMOTE : Synthetic Minority Over-Sampling
Positive
Negative
Feature Space
![Page 20: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/20.jpg)
II. How to solve Class Imbalance problem
Over Sampling
① How to create repeatedly instances of positive class.
② SMOTE : Synthetic Minority Over-Sampling
Positive
Negative
Feature Space
![Page 21: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/21.jpg)
II. How to solve Class Imbalance problem
Over Sampling
② SMOTE : Synthetic Minority Over-Sampling.
Positive
Negative
Select a observation from the minority(positive) class.
Feature Space
![Page 22: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/22.jpg)
II. How to solve Class Imbalance problem
Over Sampling
② SMOTE : Synthetic Minority Over-Sampling.
Positive
Negative
Select 𝑘 nearst neighbors.(𝑘 = ℎ𝑦𝑝𝑒𝑟𝑝𝑎𝑟𝑎𝑚𝑒𝑡𝑒𝑟)
𝑘 = 3
Feature Space
![Page 23: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/23.jpg)
II. How to solve Class Imbalance problem
Over Sampling
② SMOTE : Synthetic Minority Over-Sampling.
Positive
Negative
Create a minority(positive) class arbitrarily in a straight line between two points.
𝑘 = 3
Feature Space
![Page 24: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/24.jpg)
II. How to solve Class Imbalance problem
Over Sampling
② SMOTE : Synthetic Minority Over-Sampling.
Positive
Negative
Create a minority(positive) class arbitrarily in a straight line between two points. -> Interpolation
𝑘 = 3
Feature Space
![Page 25: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/25.jpg)
II. How to solve Class Imbalance problem
Over Sampling
② SMOTE : Synthetic Minority Over-Sampling.
Positive
Negative
𝑘 = 3
Feature Space
Create a minority(positive) class arbitrarily in a straight line between two points. -> Interpolation
![Page 26: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/26.jpg)
II. How to solve Class Imbalance problem
Over Sampling
② SMOTE : Synthetic Minority Over-Sampling.
After generation is complete, apply a classification algorithm.
Feature Space
Positive
Negative
Generated
![Page 27: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/27.jpg)
II. How to solve Class Imbalance problem
Under Sampling
① RUS : Random Under Sampling.
Positive
Negative
Feature Space
![Page 28: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/28.jpg)
II. How to solve Class Imbalance problem
Under Sampling
① RUS : Random Under Sampling.
Positive
Negative
Remove negative(majority) observations randomly.
Feature Space
![Page 29: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/29.jpg)
II. How to solve Class Imbalance problem
Under Sampling
① RUS : Random Under Sampling.
Positive
Negative
Remove negative(majority) observations randomly.
Remove
Feature Space
![Page 30: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/30.jpg)
II. How to solve Class Imbalance problem
Under Sampling
① RUS : Random Under Sampling.
Positive
Negative
Remove negative(majority) observations randomly.
Feature Space
![Page 31: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/31.jpg)
II. How to solve Class Imbalance problem
Comparison Under sampling with Over sampling
Advantages Disadvantages
Under Sampling
① We reduce the size of the training dataset by removing the data from the negative (majority) class.
② Time to train model when using under sampling techniques is shorter than oversampling techniques.
① Because we remove observations, we can not use the information that we have in the modeling process.
Over Sampling
① Since observations are not removed, no loss of information occurs.
② Because of the use of interpolation, class boundaries do not change. That is, the distribution of the positive (minority) class does not change.
① Because it creates observations for the positive (minority) class, it takes larger time to train the training data than undersampling.
![Page 32: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/32.jpg)
II. How to solve Class Imbalance problem
Sampling Techniques
Algorithm Techniques
Feature selection Techniques
Over Sampling vs. Under Sampling
Boosting, AdaBoost,……
Lots of feature selection techniques
![Page 33: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/33.jpg)
II. How to solve Class Imbalance problem
Boosting
Boosting is an ensemble method that creates a predictive model by continuously building weak models to better classify misclassified observations.
![Page 34: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/34.jpg)
II. How to solve Class Imbalance problem
Boosting
Boosting is an ensemble method that creates a predictive model by continuously building weak models to better classify misclassified observations.
It takes a long time to generate a weak classifier based on misclassified observations, but it performs better than normal classifiers(ex. Decision Tree, logistic regression).
![Page 35: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/35.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
![Page 36: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/36.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
Feature Space
![Page 37: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/37.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
①
Feature Space
![Page 38: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/38.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
①
Positive
Negative
②
Feature Space
![Page 39: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/39.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
②
③
①
Feature Space
![Page 40: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/40.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
②
③④
①
Feature Space
![Page 41: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/41.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
②
③④
⑤
①
Feature Space
![Page 42: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/42.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
②
③④
⑤
⑥
①
Feature Space
![Page 43: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/43.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
②
③④
⑤
⑥
①
Feature Space
![Page 44: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/44.jpg)
II. How to solve Class Imbalance problem
AdaBoost(Adaptive Boosting)
AdaBoost is a boosting method that creates weak classifiers while giving larger weight to misclassified observations than well-classified observations.
Positive
Negative
Feature Space
![Page 45: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/45.jpg)
II. How to solve Class Imbalance problem
![Page 46: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/46.jpg)
Contents
I. Introduction to Class Imbalance problem
II. How to solve Class Imbalance problem
III. RUSBoost vs. SMOTEBoost
IV.Result of experiments
V. Conclusion
![Page 47: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/47.jpg)
RUSBoost : A Hybrid Approach to Alleviating Class Imbalance
III. RUSBoost vs. SMOTEBoost
IEEE TSMC(IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS)
![Page 48: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/48.jpg)
RUSBoost : A Hybrid Approach to Alleviating Class Imbalance
III. RUSBoost vs. SMOTEBoost
![Page 49: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/49.jpg)
RUSBoost : A Hybrid Approach to Alleviating Class Imbalance
III. RUSBoost vs. SMOTEBoost
Hybrid = (Sampling + Algorithm) technique
![Page 50: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/50.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 51: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/51.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
①
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 52: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/52.jpg)
Positive
Negative
Removed
①
② RUSBoost(RUS + AdaBoost)
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 53: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/53.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
①
②
③
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 54: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/54.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
①
②
③
④
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 55: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/55.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
①
②
③
④
⑤
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 56: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/56.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
①
②
③
④
⑤
⑥
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 57: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/57.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
①
②
③
④
⑤
⑦
⑥
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 58: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/58.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
①
②
③
④
⑤
⑧
⑦
⑥
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 59: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/59.jpg)
RUSBoost(RUS + AdaBoost)
Positive
Negative
Removed
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 60: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/60.jpg)
III. RUSBoost vs. SMOTEBoost
RUSBoost Pesudo code
![Page 61: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/61.jpg)
III. RUSBoost vs. SMOTEBoost
RUSBoost Pseudo code
![Page 62: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/62.jpg)
SMOTEBoost : Improving Prediction of the Minority Class in Boosting
III. RUSBoost vs. SMOTEBoost
European Conference on Principles of Data Mining and Knowledge Discovery - 2003
![Page 63: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/63.jpg)
Positive
Negative
SMOTEBoost(SMOTE + AdaBoost)
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 64: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/64.jpg)
SMOTEBoost(SMOTE + AdaBoost)
Positive
Negative
Generated
①
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 65: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/65.jpg)
SMOTEBoost(SMOTE + AdaBoost)
Positive
Negative
Generated
①
②
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 66: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/66.jpg)
SMOTEBoost(SMOTE + AdaBoost)
Positive
Negative
Generated
①
②
③
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 67: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/67.jpg)
SMOTEBoost(SMOTE + AdaBoost)
Positive
Negative
Generated
①
②
③
④
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 68: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/68.jpg)
SMOTEBoost(SMOTE + AdaBoost)
Positive
Negative
Generated
①
②
③
④
⑤
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 69: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/69.jpg)
SMOTEBoost(SMOTE + AdaBoost)
Positive
Negative
Generated
①
②
③
④
⑤
⑥ ⑦
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 70: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/70.jpg)
SMOTEBoost(SMOTE + AdaBoost)
Positive
Negative
Generated
①
②
③
④
⑤
⑥⑧
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 71: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/71.jpg)
Positive
Negative
SMOTEBoost(SMOTE + AdaBoost)
III. RUSBoost vs. SMOTEBoost
Feature Space
![Page 72: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/72.jpg)
III. RUSBoost vs. SMOTEBoost
SMOTEBoost Pseudo code
![Page 73: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/73.jpg)
III. RUSBoost vs. SMOTEBoost
SMOTEBoost Pseudo code
![Page 74: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/74.jpg)
Contents
I. Introduction to Class Imbalance problem
II. How to solve Class Imbalance problem
III. RUSBoost vs. SMOTEBoost
IV.Result of experiments
V. Conclusion
![Page 75: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/75.jpg)
IV. Result of experiments
Datasets
Experiments were conducted on 15 class imbalance data.
To ensure independence of each result value, 10-fold cross validation was performed 10 times in total.
![Page 76: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/76.jpg)
IV. Result of experiments
Multiple Comparison(다중비교)
After the analysis of variance , this hypothesis test is conducted when the null hypothesis is rejected.
![Page 77: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/77.jpg)
IV. Result of experiments
Multiple Comparison(다중비교)
After the analysis of variance , this hypothesis test is conducted when the null hypothesis is rejected.
Tests are conducted by grouping the two groups together and testing how similar the groups are.
Group1 Group2 Group3 Group4
![Page 78: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/78.jpg)
IV. Result of experiments
Multiple Comparison(다중비교)
Group1 Group2 Group3 Group4
After the analysis of variance , this hypothesis test is conducted when the null hypothesis is rejected.
Tests are conducted by grouping the two groups together and testing how similar the groups are.
![Page 79: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/79.jpg)
IV. Result of experiments
Multiple Comparison(다중비교)
Group1 Group2 Group3 Group4
After the analysis of variance , this hypothesis test is conducted when the null hypothesis is rejected.
Tests are conducted by grouping the two groups together and testing how similar the groups are.
![Page 80: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/80.jpg)
IV. Result of experiments
Result
![Page 81: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/81.jpg)
Contents
I. Introduction to Class Imbalance problem
II. How to solve Class Imbalance problem
III. RUSBoost vs. SMOTEBoost
IV.Result of experiments
V. Conclusion
![Page 82: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/82.jpg)
V. Conclusion
Conclusion
You should use the appropriate algorithm for your problem situation.
For example, if you do not know whether the data is very large and can be operated on the memory, it is recommended to select the RUS algorithms.
If the training dataset is very small and the number of positive (minority) class is also small, you should use the SMOTE algorithms.
![Page 83: RUSBoost : A Hybrid Approach to Alleviating Class Imbalancedmqa.korea.ac.kr/uploads/seminar/190517 DMQA Open Seminar... · 2019-05-21 · RUSBoost : A Hybrid Approach to Alleviating](https://reader033.vdocument.in/reader033/viewer/2022042414/5f2f1316cef5c516df475264/html5/thumbnails/83.jpg)
Thank you.