Supervised Distance Preserving Projections (SDPP)

SDPP is a supervised dimensionality reduction technique that finds a subset of vectors that can best reproduce changes in a response space. The method can be solved by using gradient descent or by solving a semidefinite program. You can download the SDPP MATLAB files from below

• sdpp.zip

Please give reference to our contribution below if you use our SDPP codes in your publishable work.

• Zhanxing Zhu, Timo Similä, Francesco Corona. Supervised Distance Preserving Projections. In Neural Processing Letters Volume 38 (2013), Issue 3 , pp 445-463 . [pdf]