Add iterated Ansatzes

[lcgraham]

No description provided.

[mpilosov]

One of the things I'll be doing as part of thesis work (months down the line after I wrap up the masters) is exploring the differences between inverting lambda_dim QoIs simultaneously and inverting sequentially using one at a time, passing the posteriors as ansatzs to the following inversion. Is this type of procedure what this issue is referring to or is it something else? Could you please clarify?

[lcgraham]

That is the type of procedure this issue is referring to.

On Jun 22, 2016, at 5:43 PM, Michael Pilosov notifications@github.com wrote:

One of the things I'll be doing as part of thesis work (months down the line after I wrap up the masters) is exploring the differences between inverting lambda_dim QoIs simultaneously and inverting sequentially using one at a time, passing the posteriors as ansatzs to the following inversion. Is this type of procedure what this issue is referring to or is it something else? Could you please clarify?

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.

[mathematicalmichael]

@eecsu: maps defined by rotations equispaced from [0, pi]. output dim is 50
https://imgur.com/a/hbt88u5

if you go from [0, 2pi], HALF the maps are "completely redundant" in theory... but look at how the probabilities get spread around towards the edges... rather than the center.
https://imgur.com/a/0qfQdSQ