Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
12
0

Year Published

2018
2018
2023
2023

Publication Types

Select...
4
2
1

Relationship

3
4

Authors

Journals

citations
Cited by 10 publications
(12 citation statements)
references
References 9 publications
0
12
0
Order By: Relevance
“…• The generated SB dataset is highly imbalanced, which will negatively impact the performance of classifiers as shown in [14]. The decision boundary of the fraud classifiers will be biased towards the normal class, which means suspicious bidders will poorly be detected.…”
Section: Discussionmentioning
confidence: 99%
“…• The generated SB dataset is highly imbalanced, which will negatively impact the performance of classifiers as shown in [14]. The decision boundary of the fraud classifiers will be biased towards the normal class, which means suspicious bidders will poorly be detected.…”
Section: Discussionmentioning
confidence: 99%
“…In fraud applications, the situation is more unfortunate as it is the minority (suspicious) class that is vital to detect since it carries the highest cost of misclassification. Furthermore, a screwed class distribution always deteriorates the predictive performance as demonstrated in [15].…”
Section: Sampling Of Sb Datasetmentioning
confidence: 96%
“…Swati Ganguly [10] focused on the dataset imbalance issue, and after preprocessing the data, the author implemented its dataset into three models that are Naïve Bayes, Neural network, and Decision tree. The author claims the results that Naïve Bayes is less sensitive than NN and Decision Tree in data quality and Decision tree is working better than other models on the rebalanced training dataset.…”
Section: Literature Reviewmentioning
confidence: 99%
“…By using a decision tree successfully achieve 98% accuracy. Sawati Ganguly et al [6] Ahmad Alzahrani [7] Farzana Anowar et al [8] Sulaf Elshaar et al [9] Swati Ganguly [10] Priyanka Gupta et al [11] Sulaf Elshaar et al [12] Yanjiao Dong et al [13] Jin Xiao et al [14] Sulaf Using the hybrid model, the least good accuracy is 70%, and by labelling and multi-dimensional preprocessing attain 94% accuracy while the proposed DFM-SB model accomplishes the accuracy of 99.63%, which is better than the previous models.…”
Section: Figure 3: Comparison Proposed Model With the Previous Modelmentioning
confidence: 99%