# Machine Learning/HMM

From Noisebridge

< Machine Learning(Difference between revisions)

Mschachter (Talk | contribs) |
ThomasLotze (Talk | contribs) (→R) |
||

(5 intermediate revisions by 3 users not shown) | |||

Line 5: | Line 5: | ||

=== Implementations === | === Implementations === | ||

− | |||

*[http://ghmm.sourceforge.net/ GHMM] (C) | *[http://ghmm.sourceforge.net/ GHMM] (C) | ||

+ | *[http://hmmer.janelia.org HMMER] (compiled C-apps for Protein (possibly speech) analysis) | ||

*[http://www.logilab.org/912/ logilab-hmm] (Python) | *[http://www.logilab.org/912/ logilab-hmm] (Python) | ||

*[http://www.cs.ubc.ca/~murphyk/Software/HMM/hmm.html HMM Toolbox] (MATLAB) | *[http://www.cs.ubc.ca/~murphyk/Software/HMM/hmm.html HMM Toolbox] (MATLAB) | ||

*[http://code.google.com/p/jahmm/ jahmm] (Java) | *[http://code.google.com/p/jahmm/ jahmm] (Java) | ||

+ | ** I found [http://www.mblondel.org/journal/2009/05/19/java-jruby-or-jython-for-scientific-computing-a-test-case-with-hidden-markov-models/ Mathieu Blondel's writeup] really helpful -- jahmm is a good package | ||

+ | |||

+ | ==== R ==== | ||

+ | * [http://cran.r-project.org/web/packages/HMM/index.html HMM]: very simple hidden markov models | ||

+ | * [http://cran.r-project.org/web/packages/hmm.discnp/index.html hmm.discnp]: allows observations of multiple runs | ||

+ | * [http://cran.r-project.org/web/packages/msm/index.html msm]: continuous time, with covariates, multiple runs | ||

+ | * Example | ||

+ | ** [[Machine Learning/HMM R Example | example]] | ||

+ | ** [[Machine_Learning/student_data.csv | student_data.csv]] |

## Latest revision as of 21:15, 4 August 2010

## Contents |

## [edit] Hidden Markov Models

### [edit] Papers/Tutorials

### [edit] Implementations

- GHMM (C)
- HMMER (compiled C-apps for Protein (possibly speech) analysis)
- logilab-hmm (Python)
- HMM Toolbox (MATLAB)
- jahmm (Java)
- I found Mathieu Blondel's writeup really helpful -- jahmm is a good package

#### [edit] R

- HMM: very simple hidden markov models
- hmm.discnp: allows observations of multiple runs
- msm: continuous time, with covariates, multiple runs
- Example