As far as I have used wavelets are used to smooth or filter the data without dimension reduction . Decomposition and reconstruction of data and then using ML is what I have worked on from what i see it does if you want to make inferences in the frequency domain and PCA helps in solving the curse of dimensionality ...
Dr. Aparna Sathya Murthy, you are right. I also used PCA+LSSVM and wavelet+LSSVM and wavelet+PCA+LSSVM. I also found that wavelets are used to smooth or filter the data without dimension reduction . Decomposition and reconstruction of data and then using ML is what I have worked on from what i see it does if you want to make inferences in the frequency domain and PCA helps in solving the curse of dimensionality ...
But I also found that PCA+LSSVM work more efficiently than wavelet+LSSVM.