From: Prediction of flight departure delays caused by weather conditions adopting data-driven approaches
Sources | Method | Datasets | Data period | Delay time (min) | Results | |
---|---|---|---|---|---|---|
Machine learning | Khaksar and Sheikholeslami [9] | Bayesian modeling, decision tree, cluster classification, random forest, and hybrid method | US and Iranian airline | US: 6 months, Iran: 16 months | 0–15, 15–60, 60+ | Accuracy more than 70% |
Al-Tabbakh et al. [11] | Decision tree, random forest, and REPTree | Egypt airline | Jan 2018 (1 month) | – | Accuracy around 80.3% | |
Ye et al. [12] | Multiple linear regression, support vector machine, extremely randomized trees, and LightGBM | Nanjing Lukou airline | Mar 1st 2017 to Feb 28th 2018 | 15+ | Accuracy of 86.53% | |
Atlioglu et al. [13] | 11 machine learning models. CART, KNN, GBM, XGB, and LGBM | Dammam King Fahd International Airport | Jan 1st 2017 to Dec 9th 2019 | 15+ | Accuracy around 82% | |
Neural network | Kim et al. [8] | LSTM, RNN | ATL, LAX, ORD, DFW, DEN, JFK, SFO, CLT, LAS, PHX | Jan 2010–Aug 2015 | 15+, 30+ | Accuracy of 90.95% |
Qu et al. [10] | CBAM-CondenseNet and SimAM-CNN-MLSTM | The Civil Aviation Administration of the China East China Regional Administration (ECRA) | Mar 2018–May 2019 | 15–60, 60–120, 120–240, 240+ | Accuracy of 89.8%, 91.36% | |
Yazdi et al. [14] | Stack denoising autoencoder- levenberg marquart model, SAE-LM, SDA | The Bureau of Transportation Statistics of United State Department of Transportation | For 5 years | 15+ | Accuracy of 96%, 86%, 89% |