I hope to present here some result in February, but I’ll expose what I’ve implemented so far :
- Isomap
- LLE
- Laplacian Eigenmaps
- Hessian Eigenmaps
- Diffusion Maps (in fact a variation of Laplacian Eigenmaps)
- Curvilinear Component Analysis (the reduction part)
- NonLinear Mapping (Sammon)
- My own technique (reduction, regression and projection)
- PCA (usual reduction, but robust projection with an a priori term)
The results I will show here are mainly reduction comparison between the techniques, knowing that each technique has a specific field of application : LLE is not made to respect the geodesic distances, Isomap, NLM and my technique are.
Matt, I can’t wait for this package. My work occasionally touches on manifold learning and I haven’t had the guts to re-write all my matlab code for python. This will be great.
I hope to be able to propose the code in the near future. Stay tuned 🙂
Thanks for sharing!
Hello.
I’d like to download your manifold learning toolbox that implements the above (Isomap, LLE, …) https://blog.audio-tk.com/2008/01/15/more-on-manifold-learning/
for Python.
Thank you.
Andrew M. Neiderer
neiderer@arl.army.mil
Hi Andrew,
It is available as is with scikits.learn 0.1. In the next release, it was deleted by the maintainer, but it is sitll usable.