Support Vector Machines with the Correlation Kernel for the Classification of Raman Spectra

The range of applications of Raman-based classification has expanded significantly, including applications in bacterial identification. The first stage in the classification of Raman spectra is commonly some form of preprocessing. This pre-processing greatly affects the accuracy of the results and introduces user bias and over-fitting effects. In this paper, we propose the use of Support Vector Machines with a novel correlation kernel. Results, obtained from the analysis of Raman spectra of bacteria, illustrate that the correlation kernel is "self-normalizing" and produces superior classification performance with minimal pre-processing, even on highly-noisy data obtained using inexpensive equipment. In addition, the performance does not degrade when applied to distinct test sets, a key feature of a clinically viable diagnostic application of Raman Spectroscopy.