issues
search
dattalab
/
pyhsmm-library-models
library models built on top of pyhsmm
0
stars
1
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
log_likelihood on LHSMM requires recomputation, even after resampling
#57
alexbw
closed
10 years ago
5
make library versions of old-style subHMMs
#56
mattjj
opened
10 years ago
0
What is taking single-core time during likelihood computations on EC2?
#55
alexbw
opened
10 years ago
0
serial log_likelihood calculation segfaults on giant datasets
#54
alexbw
opened
10 years ago
1
split/merge moves
#53
mattjj
opened
10 years ago
0
temperature for subHMM models
#52
mattjj
opened
10 years ago
0
don't recompute likelihoods when data division changes
#51
mattjj
closed
10 years ago
4
Memory surge (in sparse transition matrix code)
#50
alexbw
closed
10 years ago
4
Flexible IPython parallel profile usage
#49
alexbw
opened
10 years ago
4
`_stateseq_norep` and others not being set up properly
#48
mattjj
closed
10 years ago
9
Unable to train SubHMMs with size larger than Nmaxsuper=65
#47
alexbw
closed
10 years ago
11
What is being learned in "catch-all" states?
#46
alexbw
closed
10 years ago
2
left-to-right library models
#45
mattjj
opened
10 years ago
0
Parallelize log_likelihood() calculation
#44
alexbw
opened
10 years ago
0
Paring down a model for efficient pickling
#43
alexbw
closed
10 years ago
18
Work out a BD-iHMM version
#42
alexbw
opened
10 years ago
0
Bias subHMMs to use fewer components
#41
alexbw
closed
10 years ago
1
Train the subHMM with 2-4x the data size
#40
alexbw
closed
10 years ago
1
constrain subHMM transitions to be local in feature space
#39
alexbw
opened
10 years ago
0
Nmaxsuper=1 doesn't work
#38
alexbw
closed
10 years ago
3
subHMMs don't learn (much)
#37
alexbw
closed
10 years ago
30
Option to use max instead of mean likelihood in FrozenMixtureDistribution
#36
alexbw
opened
10 years ago
5
Remove load-balanced view from parallel code
#35
alexbw
closed
10 years ago
1
'zero probability under model' error happens sometimes
#34
mattjj
opened
10 years ago
5
test saving subhmm library model to disk
#33
mattjj
closed
10 years ago
2
Parallel sampling in library_subhmm_models fails
#32
alexbw
closed
10 years ago
25
Order of labels returned from parallel resampling
#31
alexbw
closed
10 years ago
4
Cannot train models on very large datasets
#30
alexbw
closed
10 years ago
0
Parallel sampling fails if likelihoods have been precomputed
#29
alexbw
closed
10 years ago
2
Sampling after unfreezing fails
#28
alexbw
opened
10 years ago
12
add and test an unfreeze() method
#27
mattjj
closed
10 years ago
2
Model truncation is broken
#26
alexbw
closed
11 years ago
3
Model reinstantiation is broken
#25
alexbw
closed
11 years ago
6
truncate HDP-based sampled model to fixed state size
#24
mattjj
closed
11 years ago
15
Reconstructing model from state sequence and data
#23
alexbw
closed
11 years ago
6
Compile failure
#22
alexbw
closed
11 years ago
6
real_data plots for HMM performing better than HSMM
#21
alexbw
opened
11 years ago
15
Tests failing on Orchestra
#20
alexbw
closed
11 years ago
7
E_step not implemented in HSMM
#19
alexbw
opened
11 years ago
3
class assertion error for duration distribution
#18
alexbw
closed
11 years ago
2
likelihoods should be computed in parallel
#17
mattjj
closed
11 years ago
1
Can't copy a sample for parallel HSMM
#16
alexbw
closed
11 years ago
3
try going to float32s for cache efficiency
#15
mattjj
closed
11 years ago
1
reduce global_model communication size
#14
mattjj
closed
11 years ago
1
parallel should only be imported once
#13
mattjj
closed
11 years ago
1
parallelize obs distn resampling
#12
mattjj
closed
11 years ago
1
make likelihood computations fast when interleaved with resampling
#11
mattjj
closed
11 years ago
1
Parallel sampling stalls
#10
alexbw
closed
11 years ago
6
Parallel sampling fails, "global_model" is undefined
#9
alexbw
closed
11 years ago
2
check on saved sample file sizes
#8
mattjj
closed
11 years ago
1
Next