ETD

• bioinformatics
• Burrows-Wheeler transform
• BWT
• eBWT
• kernel
• machine learning
• ML
• support vector machines
• SVM

This study aims to develop and evaluate novel string kernels based on the extended
Burrows-Wheeler transform (eBWT). We evaluated the performance of these methods
using support vector machines (SVM) machine learning models trained on multiple
biological datasets compared against other well-established kernels. Results show that
eBWT kernels perform especially well on long sequences.