Determining the Landing Error Scoring System after a Jump by Artificial Intelligence

Abstract

Objective: The study aims to examine the predictability of the Landing Error Scoring System (LESS) results after the jump with the Adaptive Boosting (AdaBoost) algorithm. Materials and Methods: A model has been developed by artificial intelligence to shorten the scoring system significantly. In the data preprocessing stage, 17 different items contained in the original dataset were reduced to 13. A total of 3790 data items were included in the dataset used in the study, and the dataset was divided into 4 different sub-datasets. AdaBoost was chosen to give the highest accuracy tested in five different machine learning used for regression. The model's reliability was evaluated by testing the proposed AdaBoost model with performance metrics. Results: The error score given by the clinician in the LESS was in the range of 0-86.6%. Recommended AdaBoost model for Sub1, Sub2, Sub3, and Sub4 respectively 98%, 87%, 88%, 89% accuracy has been achieved. Conclusions: The score given to the LESS's 8th, 10th, 16th, and 17th items can be predicted with high accuracy, and the total score can be reached through the model proposed in the research.