Table 8 Comparison of the accuracy value between of SDA-LM, SAE-LM, SDA and other methods
From: Flight delay prediction based on deep learning and Levenberg-Marquart algorithm
Method | Accuracy | |
|---|---|---|
Stack Denoising Autoencoder- Levenberg Marquart | Balanced dataset | Imbalanced dataset |
96% | 92% | |
Random Forest [149] | 90.2% | – |
Stack Autoencoder [151] | – | 90% |
Stack Denoising Autoencoder | 89% | 80% |
Recurrent Neural Network [24] | – | 88% |
Stack Autoencoder- Levenberg Marquart | 86% | 83% |
Supervised Learning Methods [128] | – | 86% |
Data mining + Machine Learning [126] | 85.73% | – |
Deep Belief Neural Network + Decision Tree [154] | 82.10% | – |
Cat-boost Model [125] | 80.44% | – |