An novel efficient method of multi-class support vector machine with weighted multiple kernel learning

<title>Abstract</title> <p>This paper proposes an All-in-One multi-class SVM (AIO-MSVM) based on weighted multiple kernel learning (AIO-MSVM-WMK) for enhanced complex data classification. It integrates adaptive weighted multiple kernel learning with sample optimization in a unified framework, prioritizing representative samples for efficiency. Experimental evaluations on benchmark datasets demonstrate that AIO-MSVM-WMK outperforms existing multi-class algorithms, achieving higher prediction accuracy and significantly reduced total computational time (encompassing both sampling and training phases) without additional preprocessing. Results validate its effectiveness and scalability for data. By combining weighted multiple kernel learning and sample optimization, this study advances multi-class SVMs for high-dimensional classification tasks.</p>