Adaptive dragonfly algorithm with neural-network classifier vs KNN or SVM (not disaggregated in source): Accuracy advantage in test case 2
The review reported a 52% accuracy advantage in Patil test case 2 (Ungraded).
What each review found
Current versions only| Finding | Effect | Studies | GRADE certainty |
|---|---|---|---|
Adaptive dragonfly algorithm with neural-network classifier vs KNN or SVM (not disaggregated in source) Accuracy advantage in test case 2 · Dental images of human beings · India, Patil et al., 2019 The review reported a 52% accuracy advantage in Patil test case 2 (Ungraded). Uses of Different Machine Learning Algorithms for Diagnosis of Dental Caries. (2022) All outcomes | Accuracy advantage 52 | 1 study · 40 participants | GRADE not reported |
Adaptive dragonfly algorithm with neural-network classifier vs KNN or SVM (not disaggregated in source) Accuracy advantage in test case 2 · Dental images of human beings · India, Patil et al., 2019 The review reported an 11.76% accuracy advantage in Patil test case 2 (Ungraded). Uses of Different Machine Learning Algorithms for Diagnosis of Dental Caries. (2022) All outcomes | Accuracy advantage 11.76 | 1 study · 40 participants | GRADE not reported |
Findings are shown side by side, never pooled across reviews.
The reviews
| Review | Searched up to | Included studies | AMSTAR 2 |
|---|---|---|---|
| Uses of Different Machine Learning Algorithms for Diagnosis of Dental Caries. 2022 All outcomes | Not specified | 17 | Not applicable |
Studies behind the findings
None of these reviews ties its included studies to this outcome in the data here. Cochrane reviews and reviews with headline outcomes only do not list their studies yet.
How reviews are matched to this question
Reviews word the same question differently. Four decision models (TypeSafe's Jev, Liquid AI's d1-ultra, SubQ Snap and Baseten's Elephant) judged each wording of the population, intervention, comparator and outcome against the others; two wordings count as the same when at least three of them agree and none clearly disagrees, and findings count as this question only when all four elements match. Timing, setting and subgroup never split a question: they are shown on each finding.