Abstract:
Averaging predictions from multiple models improves machine-learning algorithm performance. Multiple model performance optimization should generalize well. Models must share generalization knowledge. This article proposes a multi-kernel mutual learning method for hyperspectral classification using transfer learning of combined mid-level features. Mid-level features are computed using PCA and three-layer homogenous superpixels. Sparse reconstructed, combined mean, and uniqueness are mid-level features. A joint sparse representation model under three-scale superpixel boundaries and regions yields sparse reconstruction. The superposed manifold ranking values of multilayer superpixels determine the uniqueness of the combined mean features and the average spectra values. Minimizing divergence yields three kernels of samples in different feature spaces for mutual learning. SVM training builds classifiers using a combined kernel to optimize sample distance measurement. The proposed method outperformed several state-of-the-art competitive algorithms based on MKL and deep learning on real hyperspectral datasets.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
