Prediction of Maneuvering Status for Aerial Vehicles using Supervised Learning Methods

TL;DR

This paper explores supervised machine learning methods to predict the maneuvering status of aerial vehicles based on their positional data, aiming to improve classification accuracy for navigation and control.

Contribution

It introduces a methodology applying multiple supervised learning algorithms to classify maneuvering status, with comparative analysis of their effectiveness.

Findings

Linear and ensemble methods show high accuracy in classification.

Boosting ensemble outperforms other models in prediction.

Metrics indicate the most suitable algorithm for maneuvering status prediction.

Abstract

Aerial Vehicles follow a guided approach based on Latitude, Longitude and Altitude. This information can be used for calculating the status of maneuvering for the aerial vehicles along the line of trajectory. This is a binary classification problem and Machine Learning can be leveraged for solving such problem. In this paper we present a methodology for deriving maneuvering status and its prediction using Linear, Distance Metric, Discriminant Analysis and Boosting Ensemble supervised learning methods. We provide various metrics along the line in the results section that give condensed comparison of the appropriate algorithm for prediction of the maneuvering status.