OA-006

Aircraft Engine Classification via Custom Machine Learning in MATLAB

Built a machine learning classifier that reached 86% accuracy sorting aircraft engine types.

Built a MATLAB pipeline to clean aircraft performance data, engineer features, normalize weighted metrics, cluster engines with K-means, and classify aircraft engine types.

Role

Sole developer

Timeline

August 2025 – December 2025

Organization

University of Florida

MATLABK-means clusteringConfusion matrixData cleaningFeature engineering3D visualization

MATLAB 3D K-means cluster visualization for aircraft engine classification — K-means cluster visualization using aircraft engine performance data, including max speed, rate of climb, and gross weight.

The pipeline

I built a MATLAB pipeline that takes raw aircraft performance data and turns it into an engine-type classifier: cleaning the data, engineering features, normalizing weighted metrics, clustering engines with K-means, and classifying aircraft engine types from the result.

Results

I evaluated the classifier with a training confusion matrix across jet, piston, and propjet classes, reaching 86.16% training accuracy, a working end-to-end classification workflow rather than a black box.

MATLAB command window showing aircraft engine classification confusion matrix and training accuracy — Training confusion matrix, 86.16% accuracy across jet, piston, and propjet classes.

Documents

Full Report