Last month Tsang-Kai emailed me his proof for the solution for KL Average for von Mises-Fisher distributions. Now we are looking for ways in which we can apply this to different real engineering problems that exist.

One interesting application of von Mises-Fisher distributions is in Quantum Information theory. One way to represent a quantum state is by a bloch sphere [Nielsen and Chuang, 2010].

However, in practice one cannot clone a Quantum State as a result of the no cloning theorem, and direct measurement of a quantum state will collapse it to one of the computational basis states of the measurement. Enter Quantum Tomography, were one tries to estimate the state of a qubit. Since the state of a qubit can be represented on a sphere, one idea that has come to the field of Quantum Information theory is to use directional statistics for the representation of Quantum States, mentioned in [Shultz, 2016]. In fact, he mentioned that we can use the von Mises-Fisher distribution to represent random pure states and that we can use the Bingham distribution to represent random mixed states. Also mentioned was the usage of matrix versions of these distributions to represent unitary operations on states. In his paper Shultz utilizes Choi matricies in conjuction with von Mises-Fisher class of distributions to Dephase noise, and Bingham type distributions to Depolarize noise. Also, some researchers suggest that von Mises-Fisher distributions are useful for preparing initial pure states (before computations occur) Wudarski et al. 2020].

Also interesting and related is the field of Quantum Consensus. Most of the current work is reviewed in Marcozzi and Mostarda, 2021]. Most of the current work seems to be on the straight up qubit representation of quantum states (without any fancy distributions involved).

One idea could be to use consensus of von Mises-Fisher distributions as a method to prepare the initial pure states of a Quantum network if we wish to have the quantum network be initialized from a state of consensus!

Next Post Previous Post