slds-lmu paper_2019_multiobjective_rfms issues

slds-lmu / paper_2019_multiobjective_rfms

High Dimensional Restrictive Federated Model Selection with multi-objective Bayesian Optimization over shifted distributions

https://github.com/smilesun/tex_2018_intellisys2019_fmoboms

2 stars 0 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

potential journals

#59 smilesun opened 5 years ago
0
debug ladder

#58 smilesun opened 5 years ago
0
refactor setHyperParWrapper

#57 smilesun opened 5 years ago
0
add row names to dataframe

#56 smilesun opened 5 years ago
0
better way to observe if tuning work

#55 smilesun opened 5 years ago
0
put prob_inputs_data inside problem

#54 smilesun opened 5 years ago
0
why do we use bootstrap of openbox and lockbox to evaluate? why not just average the loss on each dataset by alpha?

#53 smilesun opened 5 years ago
0
weigh loss by datasize

#52 smilesun closed 5 years ago
1
log current multiobjective pareto in each iteration

#51 smilesun opened 5 years ago
0
publish the problem setting alone as an arxiv paper first

#50 smilesun opened 5 years ago
0
another non-reproducible error

#49 smilesun opened 5 years ago
2
test_bt_oml remove constant features problem

#48 smilesun opened 5 years ago
1
can we claim this fmo-ms can be used as general **resampling & model selection** method? (cluster arbitrary dataset and run fmo-ms)

#47 smilesun opened 5 years ago
0
mbo combined with ladder met matrix inversion problem "Error in t.default(T) : argument is not a matrix"

#46 smilesun opened 5 years ago
7
run stratified dataset partition

#45 smilesun opened 5 years ago
0
lack of datasets for evaluation of our method

#44 smilesun closed 5 years ago
0
oml 14966 500 columns got removed in removeConstant features

#43 smilesun opened 5 years ago
1
oml34536, 9981, is multiclass?

#42 smilesun closed 5 years ago
1
lido can not run jobs

#41 smilesun closed 5 years ago
1
account di36tij job error after loading packages (log stoped, not error message found)

#40 smilesun closed 5 years ago
1
oml tid 9950 has zero factor after the pca+ balanced clustering

#39 smilesun closed 5 years ago
1
lido can not download oml

#38 smilesun closed 5 years ago
2
is ksvm random?

#37 smilesun opened 5 years ago
7
strange prediction object

#36 smilesun opened 5 years ago
0
implement non-mbo model selection

#35 smilesun opened 5 years ago
0
implement random search

#34 smilesun closed 5 years ago
0
log hyper-parameter change in gperf_env

#33 smilesun closed 5 years ago
1
simulated data

#32 smilesun opened 5 years ago
0
understanding the eaf plot

#31 smilesun opened 5 years ago
0
graph model: can we eliminate the arrow from theta to data?

#30 smilesun opened 5 years ago
0
code for 3d hyper-volume calculation and aggregation

#29 smilesun opened 5 years ago
3
grid search for fso alpha hyper-parameter

#28 smilesun closed 5 years ago
0
plot of pca eigenvalues and figure out a way to choose the ratio of variance to reserve

#27 smilesun closed 5 years ago
2
we need a time evolving plot of the mmce for each data site along the mbo process (data already available)

#26 smilesun opened 5 years ago
0
t-sne visualization of high dimensional clusters

#25 smilesun opened 5 years ago
5
attack point: we need some argument to explain why we split the data into 5 parts and use 3 of them as curator dataset.

#24 smilesun opened 5 years ago
0
summary statistic or graph to the 5 data sites

#23 smilesun opened 5 years ago
1
remove openml dependencies from batchtools running

#22 smilesun opened 5 years ago
0
Refactor pre

#21 pfistfl closed 5 years ago
0
effect of extremely small size part in dataset splitting

#20 smilesun opened 5 years ago
2
effect of removeConstantFeaturesWrapper

#19 smilesun closed 5 years ago
2
tuning threshold VS tuning AUC

#18 smilesun closed 5 years ago
2
code review

#17 smilesun opened 5 years ago
0
is it possible to have: when train on the whole dataset, mmce is 1, but the cv mmce score on the same can be 0.4?

#16 smilesun closed 5 years ago
4
brier_score vs scaled brier_score as tuning criteria

#15 smilesun closed 5 years ago
4
how to choose the hyper-parameter of thresholdout-brier-score

#14 smilesun opened 5 years ago
0
shall we ever use thresholdout-auc-feat-sel to compare with our method?

#13 smilesun opened 5 years ago
2
which measure should we use to compare the final result? auc, mmce or ?

#12 smilesun opened 5 years ago
0
R implementaion of thresholdout auc combined with MBO

#11 smilesun closed 5 years ago
0
removeConstanceFeatureWrapper Does not work for our code

#10 smilesun closed 5 years ago
1