Faithful Stochastic Proximity Embedding (FSPE) is a new non-linear dimensionality reduction (DR) method, which is called to visualize different data sets, such as hyperspectral images. FSPE overcomes the main shortcomings of the DR by sending the false neighbour points away, and preserving the neighbourhood relation to the true neighbours points, which are inside the local neighbourhood. The visualization of our proposed method displays the faithful, useful and meaningful colours, where the objects of the image can be easily distinguished.
- S.A. Najim, I.S. Lim, P. Wittek and M. Jones. FSPE: Visualisation of Hyperspectral Imagery Using Faithful Stochastic Proximity Embedding. IEEE Geoscience and Remote Science Letters. 12(1), pp:18-22, 2015. pdf
Dr. Safa A. Najim
Computer Science Dept.,
The FSPE's codes can be downloaded: