ThrunGroup FastForest issues

ThrunGroup / FastForest

7 stars 0 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Added 3 experiments for Log Splitting

#245 srijaladi opened 1 year ago
0
Figure out why sample complexity is logarithmic in regression but wall clock time is still linear

#244 motiwari opened 1 year ago
0
Known issues with `repro_script.sh`

#243 motiwari opened 1 year ago
0
Random baseline

#242 motiwari closed 1 year ago
0
Needed to filter `nans` in `solve_mab`

#241 motiwari opened 1 year ago
1
Slight issue in MNIST experiments

#240 motiwari opened 1 year ago
0
Can delete this line

#239 motiwari opened 1 year ago
0
Appendix Fig 1 "number of histogram insertions queried by MABSplit" should be "performed by MABSplit"

#238 motiwari opened 1 year ago
0
Consider changing vectorization in lines 173 and 176 of histogram.py? Or using np.var calls more intelligently? (low pri)

#237 motiwari opened 1 year ago
0
Final submission task

#236 motiwari opened 1 year ago
0
Reproduce experiments

#235 vxbrandon closed 1 year ago
0
GPU dataset is in gitignore

#234 vxbrandon opened 2 years ago
0
Reproduce experiments

#233 vxbrandon closed 2 years ago
0
Models are not able to fit the flight delay dataset

#232 motiwari closed 2 years ago
5
self.labels for self.new_targets for HistogramRandomForestClassifier

#231 motiwari opened 2 years ago
0
Remove upper limit of `n_estimators=100` in budget experiments

#230 motiwari opened 2 years ago
0
Links in `experiments/datasets/data_loader.py` to flight data is incorrect

#229 motiwari opened 2 years ago
1
Minor issue in variance calculation

#228 vxbrandon opened 2 years ago
0
Remove subtraction of `impurity_curr` from computation of impurity reductions

#227 motiwari opened 2 years ago
1
Merging `new_scaling_exps` into `dataset_experiments` to unify Mo and Jey's code

#226 motiwari closed 2 years ago
0
Add ability to predict from a root node even if it hasn't been split

#225 motiwari opened 2 years ago
0
Rename `dir` variables in `produce_tables.py`

#224 motiwari opened 2 years ago
0
Standardize `max_depth=3` in all experiments

#223 motiwari opened 2 years ago
0
Revert changes to `solve_mab`

#222 motiwari opened 2 years ago
0
Recreating submission experiments on `main`

#221 motiwari closed 1 year ago
0
Dataset experiments

#220 motiwari opened 2 years ago
0
Creating PR to understand differences in experiments between current code and NeurIPS submission

#219 motiwari closed 1 year ago
1
Three tasks for rebuttal

#218 vxbrandon closed 2 years ago
0
Create boolean flag to turn on/off optimizations like vectorization

#217 motiwari closed 2 years ago
1
In scaling experiments, should measure sample complexity in addition to runtime

#216 motiwari opened 2 years ago
0
Need to increase budget in `HRPR` budget exps; can't reproduce Table 4

#215 motiwari closed 2 years ago
0
Two questions about Table 3

#214 motiwari opened 2 years ago
0
Gains have vanished in runtime experiments

#213 motiwari opened 2 years ago
0
Make sure all uncertainties are computed correctly in Tables

#212 motiwari opened 2 years ago
0
Vectorization

#211 motiwari closed 2 years ago
0
Bugs with `./repro_script.sh`

#210 motiwari closed 2 years ago
1
Fix division by 0 errors

#209 motiwari closed 2 years ago
1
Demonstrate log scaling in both classification and regression (figure out why some scaling is linear instead of logarithmic)

#208 motiwari closed 2 years ago
1
Refactor, cleanup, adding environment.yml. Tests pass

#207 motiwari closed 2 years ago
0
Solve linearity of fitting HRFC into MNIST dataset

#206 vxbrandon opened 2 years ago
0
Address discussion

#205 motiwari closed 2 years ago
1
Note to Mo: Removing binning type `RANDOM` for ERFs

#204 motiwari opened 2 years ago
0
Produce tables

#203 vxbrandon closed 2 years ago
7
Bias variance decomposition analysis

#202 vxbrandon closed 2 years ago
2
Vectorization of impurity reduction calculation and insertion in histograms

#201 vxbrandon closed 2 years ago
1
Can reduce memory usage by passing indices of child data instead of child data itself

#200 vxbrandon closed 2 years ago
0
Possible to obviate allocating entire dataset to left/right nodes and only use them on the fly?

#199 motiwari opened 2 years ago
0
How to use line profiler

#198 vxbrandon opened 2 years ago
0
Produce/Reproduce tables 1-6 and figure1

#197 vxbrandon closed 2 years ago
3
Fixes #195

#196 vxbrandon closed 2 years ago
0