<html><head><meta http-equiv="Content-Type" content="text/html charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;"><blockquote type="cite"><pre style="white-space: pre-wrap; widows: 1;">I am currently working on
<a href="http://hub.darcs.net/thielema/hmm-hmatrix">http://hub.darcs.net/thielema/hmm-hmatrix</a>
It does not need log-numbers because it normalizes all temporary results.
This way I can use fast hmatrix operations. Would normalization also be a
solution for other probabilistic models?</pre></blockquote><div>I assume these are models with finite state spaces? FYI I uploaded a Kalman filter <a href="https://github.com/idontgetoutmuch/Kalman">https://github.com/idontgetoutmuch/Kalman</a> which is also a HMM. I feel we should be able to abstract over the Markov transition kernel to describe the model without caring whether the state space is continuous or discrete but that it is about as much thinking I have done on the subject.</div><div><br></div><div apple-content-edited="true">
<div>Dominic Steinitz</div><div><a href="mailto:dominic@steinitz.org">dominic@steinitz.org</a></div><div><a href="http://idontgetoutmuch.wordpress.com">http://idontgetoutmuch.wordpress.com</a></div>
</div>
<br></body></html>