This article first uses the PCA algorithm to reduce the dimensionality of expert-based flight parameter-based flight technical assessment report data. Then, the reduced data is respectively introduced into eight machine learning algorithms, including Logistic Regression, Decision Tree, etc. for pilot qualification assessment and prediction. Next, the SMOTEE algorithm is used to optimize the above eight models respectively. Finally, this article a dopts the PCA-SMOTE-Random Forest model with the best prediction effect to evaluate and predict the flight quality. Through confusion matrix diagrams and ROC curve diagrams, the accuracy of the algorithm’s prediction is nearly 100%. Finally, this article discusses a flight technology evaluation method based on flight parameter.
Flight Technology Assessment Based on Machine Learning and SMOTE Algorithm
15.12.2023
792196 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch