Amari has worked
in the theory of estimation in statistics, and has written the notable
book, "Differential Geometric Methods in Statistics", Lecture Notes in
Statistics, Vol **28**, 1985. He has also noted that similar ideas occur in
the dynamics of neural nets and disordered systems.
Working independently, my ideas on statistical
dynamics took a similar route, but lacked any geometrical flavour.
This theory was developed in my book. It was not until I met Amari in the 1995
conference at Madeira that I realised that the correct way forward was
through information geometry.

Go to my **HOME PAGE** for more links.

© by Ray Streater, 3/8/00.