issues
search
cmaclell
/
concept_formation
Python implementations of TRESTLE, COBWEB/3, and COBWEB
MIT License
61
stars
18
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
I updated cobweb to use relative CU to reduce time complexity of the cobweb...
#54
cmaclell
closed
7 years ago
0
Update viz_logic.js
#53
ThomasHoppe
closed
7 years ago
0
Incomplete installation?
#52
ThomasHoppe
closed
5 years ago
11
readthedocs examples are broken again
#51
eharpste
opened
7 years ago
2
Consider constraining possible structure matches
#50
cmaclell
opened
7 years ago
0
Consider modifying visualize_cluster to do the clustering for you.
#49
cmaclell
closed
7 years ago
1
Visualization update
#48
eharpste
closed
7 years ago
3
Update internal sm
#47
cmaclell
closed
7 years ago
0
Structure map internally broken after updating preprocessors.
#46
cmaclell
closed
7 years ago
0
Allow Structure Mapper to accept initial mappings
#45
cmaclell
closed
7 years ago
0
fixed preprocessors
#44
cmaclell
closed
7 years ago
0
Problems with Flattening with Subcomponents
#43
eharpste
closed
7 years ago
1
Seed Examples
#42
eharpste
closed
7 years ago
1
Should instances be structure mapped to the resulting trestle concept before making predictions?
#41
cmaclell
opened
7 years ago
0
write some kind of test for infer_missing
#40
cmaclell
closed
7 years ago
1
Consider removing low probability attribute values from nodes to speed up calculations (less attribute values to compute over).
#39
cmaclell
opened
7 years ago
0
Adds support for Trestle to structure map internally.
#38
cmaclell
closed
7 years ago
2
Updated structure mapper to rely on expected correct guesses.
#37
cmaclell
closed
7 years ago
1
Attribute filter
#36
eharpste
closed
7 years ago
0
Add test cases / example for infer_missing
#35
cmaclell
opened
7 years ago
0
Updated missing
#34
cmaclell
closed
7 years ago
0
Normalizing by attributes that are hidden
#33
eharpste
closed
7 years ago
0
Normalize ec by attr count
#32
cmaclell
closed
7 years ago
0
Is attr_val_guess_gain properly handling None probabilities?
#31
cmaclell
closed
5 years ago
2
Add normalization of expected correct guesses by number of attributes
#30
cmaclell
closed
7 years ago
0
Mixed numeric nominal attr
#29
eharpste
closed
7 years ago
0
Bic cluster
#28
eharpste
closed
7 years ago
0
Potentially a problem normalizing by inner most attribute
#27
eharpste
opened
7 years ago
0
Normalize by attr name
#26
cmaclell
closed
7 years ago
0
Make it possible to have either nominal or numeric values on an attribute
#25
cmaclell
closed
7 years ago
0
Look at normalizing continuous attribute values by general attr name
#24
cmaclell
closed
7 years ago
0
0.3 docs and tests
#23
cmaclell
closed
7 years ago
0
Update structure map search
#22
cmaclell
closed
7 years ago
0
Update Documentation Throughout
#21
eharpste
closed
7 years ago
2
Are missing attributes handled correctly?
#20
cmaclell
closed
7 years ago
0
Visualization
#19
eharpste
closed
8 years ago
0
Ensure standardize apart is working correctly and check if performance can be improved.
#18
cmaclell
opened
8 years ago
1
Check that Structure mapper properly handles conflicting nominal numeric values for an attribute.
#17
cmaclell
closed
7 years ago
2
Make structure mapping faster
#16
cmaclell
closed
7 years ago
2
Heuristic in structure mapper not as good as expected
#15
cmaclell
closed
7 years ago
1
Modified structure mapper to accept instances or concept av tables.
#14
cmaclell
closed
8 years ago
0
Fix indexing errors
#13
eharpste
closed
8 years ago
0
Preprocessor Pipelines can be broken by one-way preprocessors
#12
eharpste
closed
7 years ago
2
Tracking past performance across all attribute values (rather than for each individual attribute value)
#11
cmaclell
closed
7 years ago
3
Structure Mapping at every node during TRESTLE categorization
#10
cmaclell
closed
5 years ago
1
Handeling Negated Literals
#9
cmaclell
opened
8 years ago
2
Numerics in relations throw indexing exception.
#8
eharpste
closed
8 years ago
2
Ensure that numerical components are handled correctly. I.e., review the probability of correctly guessing an attribute under a normal distribution. Then have acuity be a prior (similar to the laplace smoothing parameter alpha).
#7
cmaclell
closed
8 years ago
0
Add ability to match variables to constants (not just other variables)
#6
cmaclell
closed
8 years ago
0
Add ability to assign component values to relations.
#5
cmaclell
closed
8 years ago
0
Previous
Next