Leogrin commited on
Commit
dabc0f5
1 Parent(s): 02c346e

update-datasets (#2)

Browse files

- update datasets to new version (008fce1c4499b73b7996f7919d03048a04acd5d8)

This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +5 -0
  2. README.md +74 -75
  3. clf_cat/KDDCup09_upselling.csv +0 -0
  4. clf_cat/albert.csv +0 -0
  5. clf_cat/compas-two-years.csv +0 -0
  6. clf_cat/compass.csv +0 -0
  7. clf_cat/covertype.csv +2 -2
  8. clf_cat/default-of-credit-card-clients.csv +0 -0
  9. clf_cat/electricity.csv +0 -0
  10. clf_cat/eye_movements.csv +0 -0
  11. clf_cat/rl.csv +0 -0
  12. clf_cat/road-safety.csv +2 -2
  13. clf_num/Bioresponse.csv +0 -0
  14. clf_num/Diabetes130US.csv +0 -0
  15. clf_num/Higgs.csv +2 -2
  16. clf_num/MagicTelescope.csv +0 -0
  17. clf_num/MiniBooNE.csv +2 -2
  18. clf_num/bank-marketing.csv +0 -0
  19. clf_num/california.csv +0 -0
  20. clf_num/covertype.csv +2 -2
  21. clf_num/credit.csv +0 -0
  22. clf_num/default-of-credit-card-clients.csv +0 -0
  23. clf_num/electricity.csv +0 -0
  24. clf_num/eye_movements.csv +0 -0
  25. clf_num/heloc.csv +0 -0
  26. clf_num/house_16H.csv +0 -0
  27. clf_num/jannis.csv +2 -2
  28. clf_num/kdd_ipums_la_97-small.csv +0 -0
  29. clf_num/phoneme.csv +0 -0
  30. clf_num/pol.csv +0 -0
  31. clf_num/wine.csv +0 -0
  32. git +0 -0
  33. reg_cat/{OnlineNewsPopularity.csv → Airlines_DepDelay_1M.csv} +2 -2
  34. reg_num/isolet.csv → reg_cat/Allstate_Claims_Severity.csv +2 -2
  35. reg_cat/Bike_Sharing_Demand.csv +0 -0
  36. reg_cat/Brazilian_houses.csv +0 -0
  37. reg_cat/Mercedes_Benz_Greener_Manufacturing.csv +0 -0
  38. reg_cat/SGEMM_GPU_kernel_performance.csv +2 -2
  39. reg_cat/abalone.csv +0 -0
  40. reg_cat/analcatdata_supreme.csv +0 -0
  41. reg_num/year.csv → reg_cat/delays_zurich_transport.csv +2 -2
  42. reg_cat/diamonds.csv +0 -0
  43. reg_cat/house_sales.csv +0 -0
  44. reg_cat/{black_friday.csv → medical_charges.csv} +0 -0
  45. reg_cat/nyc-taxi-green-dec-2016.csv +2 -2
  46. reg_cat/particulate-matter-ukair-2017.csv +2 -2
  47. reg_cat/seattlecrime6.csv +0 -0
  48. reg_cat/topo_2_1.csv +3 -0
  49. reg_cat/visualizing_soil.csv +0 -0
  50. reg_cat/yprop_4_1.csv +0 -0
.gitattributes CHANGED
@@ -65,3 +65,8 @@ reg_cat/nyc-taxi-green-dec-2016.csv filter=lfs diff=lfs merge=lfs -text
65
  reg_cat/particulate-matter-ukair-2017.csv filter=lfs diff=lfs merge=lfs -text
66
  clf_cat/road-safety.csv filter=lfs diff=lfs merge=lfs -text
67
  clf_cat/covertype.csv filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
65
  reg_cat/particulate-matter-ukair-2017.csv filter=lfs diff=lfs merge=lfs -text
66
  clf_cat/road-safety.csv filter=lfs diff=lfs merge=lfs -text
67
  clf_cat/covertype.csv filter=lfs diff=lfs merge=lfs -text
68
+ reg_num/delays_zurich_transport.csv filter=lfs diff=lfs merge=lfs -text
69
+ reg_cat/Allstate_Claims_Severity.csv filter=lfs diff=lfs merge=lfs -text
70
+ reg_cat/Airlines_DepDelay_1M.csv filter=lfs diff=lfs merge=lfs -text
71
+ reg_cat/delays_zurich_transport.csv filter=lfs diff=lfs merge=lfs -text
72
+ reg_cat/topo_2_1.csv filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -81,14 +81,11 @@ subtle, but we try to keep simulated datasets if learning these datasets are of
81
  - **Not too small**. We remove datasets with too few features (< 4) and too few samples (< 3 000). For
82
  benchmarks on numerical features only, we remove categorical features before checking if enough
83
  features and samples are remaining.
84
- - **Not too easy**. We remove datasets which are too easy. Specifically, we remove a dataset if a default
85
- Logistic Regression (or Linear Regression for regression) reach a score whose relative difference
86
- with the score of both a default Resnet (from Gorishniy et al. [2021]) and a default HistGradientBoosting model (from scikit learn) is below 5%. Other benchmarks use different metrics to
87
- remove too easy datasets, like removing datasets which can be learnt perfectly by a single decision
88
- classifier [Bischl et al., 2021], but this does not account for different Bayes rate of different datasets.
89
- As tree-based methods have been shown to be superior to Logistic Regression [Fernández-Delgado
90
- et al., 2014] in our setting, a close score for these two types of models indicates that we might
91
- already be close to the best achievable score.
92
  - **Not deterministic**. We remove datasets where the target is a deterministic function of the data. This
93
  mostly means removing datasets on games like poker and chess. Indeed, we believe that these
94
  datasets are very different from most real-world tabular datasets, and should be studied separately
@@ -97,79 +94,81 @@ datasets are very different from most real-world tabular datasets, and should be
97
 
98
 
99
  **Numerical Classification**
100
- |dataset_name| n_samples| n_features| original_link| new_link|
101
- |----|----|----|----|----|
102
- |credit| 16714| 10 |https://openml.org/d/151 |https://www.openml.org/d/44089|
103
- |california |20634 |8 |https://openml.org/d/293 |https://www.openml.org/d/44090|
104
- |wine |2554 |11 |https://openml.org/d/722 |https://www.openml.org/d/44091|
105
- |electricity| 38474 |7| https://openml.org/d/821 |https://www.openml.org/d/44120|
106
- |covertype |566602 |10 |https://openml.org/d/993| https://www.openml.org/d/44121|
107
- |pol |10082 |26 |https://openml.org/d/1120 |https://www.openml.org/d/44122|
108
- |house_16H |13488| 16 |https://openml.org/d/1461| https://www.openml.org/d/44123|
109
- |kdd_ipums_la_97-small| 5188 |20 |https://openml.org/d/1489 |https://www.openml.org/d/44124|
110
- |MagicTelescope| 13376| 10| https://openml.org/d/41150 |https://www.openml.org/d/44125|
111
- |bank-marketing |10578 |7 |https://openml.org/d/42769| https://www.openml.org/d/44126|
112
- |phoneme |3172| 5 |https://openml.org/d/1044| https://www.openml.org/d/44127|
113
- |MiniBooNE| 72998| 50 |https://openml.org/d/41168 |https://www.openml.org/d/44128|
114
- |Higgs| 940160 |24| https://www.kaggle.com/c/GiveMeSomeCredit/data?select=cs-training.csv |https://www.openml.org/d/44129|
115
- |eye_movements| 7608 |20 |https://www.dcc.fc.up.pt/ltorgo/Regression/cal_housing.html |https://www.openml.org/d/44130|
116
- |jannis |57580 |54 |https://archive.ics.uci.edu/ml/datasets/wine+quality |https://www.openml.org/d/44131|
 
117
 
118
- **Categorical Classification**
119
- |dataset_name |n_samples| n_features |original_link |new_link|
120
- |----|----|----|----|----|
121
- |electricity |38474| 8 |https://openml.org/d/151| https://www.openml.org/d/44156|
122
- |eye_movements |7608 |23| https://openml.org/d/1044 |https://www.openml.org/d/44157|
123
- |covertype |423680| 54| https://openml.org/d/1114 |https://www.openml.org/d/44159|
124
- |rl |4970 |12 |https://openml.org/d/1596 |https://www.openml.org/d/44160|
125
- |road-safety| 111762 |32 |https://openml.org/d/41160 |https://www.openml.org/d/44161|
126
- |compass |16644 |17 |https://openml.org/d/42803 |https://www.openml.org/d/44162|
127
- |KDDCup09_upselling |5128 |49 |https://www.kaggle.com/datasets/danofer/compass?select=cox-violent-parsed.csv |https://www.openml.org/d/44186|
128
-
129
- **Numerical Regression**
130
- |dataset_name| n_samples| n_features| original_link| new_link|
131
- |----|----|----|----|----|
132
- |cpu_act |8192 |21| https://openml.org/d/197 |https://www.openml.org/d/44132|
133
- |pol | 15000| 26 |https://openml.org/d/201| https://www.openml.org/d/44133|
134
- |elevators |16599 |16 |https://openml.org/d/216| https://www.openml.org/d/44134|
135
- |isolet |7797| 613| https://openml.org/d/300| https://www.openml.org/d/44135|
136
- |wine_quality |6497 |11| https://openml.org/d/287 | https://www.openml.org/d/44136|
137
- |Ailerons |13750 |33| https://openml.org/d/296 | https://www.openml.org/d/44137|
138
- |houses |20640| 8| https://openml.org/d/537 | https://www.openml.org/d/44138|
139
- |house_16H |22784| 16 |https://openml.org/d/574 | https://www.openml.org/d/44139|
140
- |diamonds |53940| 6| https://openml.org/d/42225 | https://www.openml.org/d/44140|
141
- |Brazilian_houses |10692| 8 |https://openml.org/d/42688 | https://www.openml.org/d/44141|
142
- |Bike_Sharing_Demand| 17379| 6| https://openml.org/d/42712 | https://www.openml.org/d/44142|
143
- |nyc-taxi-green-dec-2016 |581835| 9| https://openml.org/d/42729 | https://www.openml.org/d/44143|
144
- |house_sales |21613 |15 | https://openml.org/d/42731| https://www.openml.org/d/44144|
145
- |sulfur |10081| 6 | https://openml.org/d/23515 | https://www.openml.org/d/44145|
146
- |medical_charges | 163065 |3 | https://openml.org/d/42720 | https://www.openml.org/d/44146|
147
- |MiamiHousing2016 |13932| 13 |https://openml.org/d/43093 | https://www.openml.org/d/44147|
148
- |superconduct |21263 |79| https://openml.org/d/43174 | https://www.openml.org/d/44148|
149
- |california |20640| 8 |https://www.dcc.fc.up.pt/ ltorgo/Regression/cal_housing.html |https://www.openml.org/d/44025|
150
- |fifa |18063 |5 |https://www.kaggle.com/datasets/stefanoleone992/fifa-22-complete-player-dataset| https://www.openml.org/d/44026|
151
- |year |515345 |90 |https://archive.ics.uci.edu/ml/datasets/yearpredictionmsd| https://www.openml.org/d/44027|
152
 
 
 
 
 
 
 
 
 
 
 
153
 
154
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
155
 
156
 
157
  **Categorical Regression**
158
- |dataset_name| n_samples| n_features| original_link| new_link|
159
- |----|----|----|----|----|
160
- |yprop_4_1 |8885 |62 |https://openml.org/d/416 |https://www.openml.org/d/44054|
161
- |analcatdata_supreme |4052| 7 |https://openml.org/d/504 |https://www.openml.org/d/44055|
162
- |visualizing_soil |8641| 4 |https://openml.org/d/688 |https://www.openml.org/d/44056|
163
- |black_friday |166821| 9 |https://openml.org/d/41540| https://www.openml.org/d/44057|
164
- |diamonds | 53940| 9| https://openml.org/d/42225| https://www.openml.org/d/44059|
165
- |Mercedes_Benz_Greener_Manufacturing |4209 |359| https://openml.org/d/42570 |https://www.openml.org/d/44061|
166
- |Brazilian_houses| 10692| 11 |https://openml.org/d/42688 |https://www.openml.org/d/44062|
167
- |Bike_Sharing_Demand| 17379| 11 |https://openml.org/d/42712 |https://www.openml.org/d/44063|
168
- |OnlineNewsPopularity |39644| 59| https://openml.org/d/42724| https://www.openml.org/d/44064|
169
- |nyc-taxi-green-dec-2016| 581835 |16 |https://openml.org/d/42729|https://www.openml.org/d/44065|
170
- |house_sales | 21613| 17| https://openml.org/d/42731| https://www.openml.org/d/44066|
171
- |particulate-matter-ukair-2017 |394299 |6| https://openml.org/d/42207| https://www.openml.org/d/44068|
172
- |SGEMM_GPU_kernel_performance | 241600| 9 |https://openml.org/d/43144| https://www.openml.org/d/44069|
 
 
 
 
173
 
174
 
175
  ### Dataset Curators
@@ -184,4 +183,4 @@ Léo Grinsztajn, Edouard Oyallon, Gaël Varoquaux.
184
 
185
  Léo Grinsztajn, Edouard Oyallon, Gaël Varoquaux. Why do tree-based models still outperform deep
186
  learning on typical tabular data?. NeurIPS 2022 Datasets and Benchmarks Track, Nov 2022, New
187
- Orleans, United States. ffhal-03723551v2f
 
81
  - **Not too small**. We remove datasets with too few features (< 4) and too few samples (< 3 000). For
82
  benchmarks on numerical features only, we remove categorical features before checking if enough
83
  features and samples are remaining.
84
+ - **Not too easy**. We remove datasets which are too easy. Specifically, we remove a dataset if a simple model (max of a single tree and a regression, logistic or OLS)
85
+ reaches a score whose relative difference with the score of both a default Resnet (from Gorishniy et al. [2021]) and a default HistGradientBoosting model (from scikit learn)
86
+ is below 5%. Other benchmarks use different metrics to remove too easy datasets, like removing datasets perfectly separated by a single decision classifier [Bischl et al., 2021],
87
+ but this ignores varying Bayes rate across datasets. As tree ensembles are superior to simple trees and logistic regresison [Fernández-Delgado et al., 2014],
88
+ a close score for the simple and powerful models suggests that we are already close to the best achievable score.
 
 
 
89
  - **Not deterministic**. We remove datasets where the target is a deterministic function of the data. This
90
  mostly means removing datasets on games like poker and chess. Indeed, we believe that these
91
  datasets are very different from most real-world tabular datasets, and should be studied separately
 
94
 
95
 
96
  **Numerical Classification**
97
+ |dataset_name|n_samples|n_features|original_link|new_link|
98
+ |---|---|---|---|---|
99
+ |electricity|38474.0|7.0|https://www.openml.org/d/151|https://www.openml.org/d/44120|
100
+ |covertype|566602.0|10.0|https://www.openml.org/d/293|https://www.openml.org/d/44121|
101
+ |pol|10082.0|26.0|https://www.openml.org/d/722|https://www.openml.org/d/44122|
102
+ |house_16H|13488.0|16.0|https://www.openml.org/d/821|https://www.openml.org/d/44123|
103
+ |MagicTelescope|13376.0|10.0|https://www.openml.org/d/1120|https://www.openml.org/d/44125|
104
+ |bank-marketing|10578.0|7.0|https://www.openml.org/d/1461|https://www.openml.org/d/44126|
105
+ |Bioresponse|3434.0|419.0|https://www.openml.org/d/4134|https://www.openml.org/d/45019|
106
+ |MiniBooNE|72998.0|50.0|https://www.openml.org/d/41150|https://www.openml.org/d/44128|
107
+ |default-of-credit-card-clients|13272.0|20.0|https://www.openml.org/d/42477|https://www.openml.org/d/45020|
108
+ |Higgs|940160.0|24.0|https://www.openml.org/d/42769|https://www.openml.org/d/44129|
109
+ |eye_movements|7608.0|20.0|https://www.openml.org/d/1044|https://www.openml.org/d/44130|
110
+ |Diabetes130US|71090.0|7.0|https://www.openml.org/d/4541|https://www.openml.org/d/45022|
111
+ |jannis|57580.0|54.0|https://www.openml.org/d/41168|https://www.openml.org/d/45021|
112
+ |heloc|10000.0|22.0|"https://www.kaggle.com/datasets/averkiyoliabev/home-equity-line-of-creditheloc?select=heloc_dataset_v1+%281%29.csv"|https://www.openml.org/d/45026|
113
+ |credit|16714.0|10.0|"https://www.kaggle.com/c/GiveMeSomeCredit/data?select=cs-training.csv"|https://www.openml.org/d/44089|
114
+ |california|20634.0|8.0|"https://www.dcc.fc.up.pt/ltorgo/Regression/cal_housing.html"|https://www.openml.org/d/45028|
115
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
116
 
117
+ **Categorical Classification**
118
+ |dataset_name|n_samples|n_features|original_link|new_link|
119
+ |---|---|---|---|---|
120
+ |electricity|38474.0|8.0|https://www.openml.org/d/151|https://www.openml.org/d/44156|
121
+ |eye_movements|7608.0|23.0|https://www.openml.org/d/1044|https://www.openml.org/d/44157|
122
+ |covertype|423680.0|54.0|https://www.openml.org/d/1596|https://www.openml.org/d/44159|
123
+ |albert|58252.0|31.0|https://www.openml.org/d/41147|https://www.openml.org/d/45035|
124
+ |compas-two-years|4966.0|11.0|https://www.openml.org/d/42192|https://www.openml.org/d/45039|
125
+ |default-of-credit-card-clients|13272.0|21.0|https://www.openml.org/d/42477|https://www.openml.org/d/45036|
126
+ |road-safety|111762.0|32.0|https://www.openml.org/d/42803|https://www.openml.org/d/45038|
127
 
128
 
129
+ **Numerical Regression**
130
+ |dataset_name|n_samples|n_features|original_link|new_link|
131
+ |---|---|---|---|---|
132
+ |cpu_act|8192.0|21.0|https://www.openml.org/d/197|https://www.openml.org/d/44132|
133
+ |pol|15000.0|26.0|https://www.openml.org/d/201|https://www.openml.org/d/44133|
134
+ |elevators|16599.0|16.0|https://www.openml.org/d/216|https://www.openml.org/d/44134|
135
+ |wine_quality|6497.0|11.0|https://www.openml.org/d/287|https://www.openml.org/d/44136|
136
+ |Ailerons|13750.0|33.0|https://www.openml.org/d/296|https://www.openml.org/d/44137|
137
+ |yprop_4_1|8885.0|42.0|https://www.openml.org/d/416|https://www.openml.org/d/45032|
138
+ |houses|20640.0|8.0|https://www.openml.org/d/537|https://www.openml.org/d/44138|
139
+ |house_16H|22784.0|16.0|https://www.openml.org/d/574|https://www.openml.org/d/44139|
140
+ |delays_zurich_transport|5465575.0|9.0|https://www.openml.org/d/40753|https://www.openml.org/d/45034|
141
+ |diamonds|53940.0|6.0|https://www.openml.org/d/42225|https://www.openml.org/d/44140|
142
+ |Brazilian_houses|10692.0|8.0|https://www.openml.org/d/42688|https://www.openml.org/d/44141|
143
+ |Bike_Sharing_Demand|17379.0|6.0|https://www.openml.org/d/42712|https://www.openml.org/d/44142|
144
+ |nyc-taxi-green-dec-2016|581835.0|9.0|https://www.openml.org/d/42729|https://www.openml.org/d/44143|
145
+ |house_sales|21613.0|15.0|https://www.openml.org/d/42731|https://www.openml.org/d/44144|
146
+ |sulfur|10081.0|6.0|https://www.openml.org/d/23515|https://www.openml.org/d/44145|
147
+ |medical_charges|163065.0|5.0|https://www.openml.org/d/42720|https://www.openml.org/d/44146|
148
+ |MiamiHousing2016|13932.0|14.0|https://www.openml.org/d/43093|https://www.openml.org/d/44147|
149
+ |superconduct|21263.0|79.0|https://www.openml.org/d/43174|https://www.openml.org/d/44148|
150
 
151
 
152
  **Categorical Regression**
153
+ |dataset_name|n_samples|n_features|original_link|new_link|
154
+ |---|---|---|---|---|
155
+ |topo_2_1|8885.0|255.0|https://www.openml.org/d/422|https://www.openml.org/d/45041|
156
+ |analcatdata_supreme|4052.0|7.0|https://www.openml.org/d/504|https://www.openml.org/d/44055|
157
+ |visualizing_soil|8641.0|4.0|https://www.openml.org/d/688|https://www.openml.org/d/44056|
158
+ |delays_zurich_transport|5465575.0|12.0|https://www.openml.org/d/40753|https://www.openml.org/d/45045|
159
+ |diamonds|53940.0|9.0|https://www.openml.org/d/42225|https://www.openml.org/d/44059|
160
+ |Allstate_Claims_Severity|188318.0|124.0|https://www.openml.org/d/42571|https://www.openml.org/d/45046|
161
+ |Mercedes_Benz_Greener_Manufacturing|4209.0|359.0|https://www.openml.org/d/42570|https://www.openml.org/d/44061|
162
+ |Brazilian_houses|10692.0|11.0|https://www.openml.org/d/42688|https://www.openml.org/d/44062|
163
+ |Bike_Sharing_Demand|17379.0|11.0|https://www.openml.org/d/42712|https://www.openml.org/d/44063|
164
+ |Airlines_DepDelay_1M|1000000.0|5.0|https://www.openml.org/d/42721|https://www.openml.org/d/45047|
165
+ |nyc-taxi-green-dec-2016|581835.0|16.0|https://www.openml.org/d/42729|https://www.openml.org/d/44065|
166
+ |abalone|4177.0|8.0|https://www.openml.org/d/42726|https://www.openml.org/d/45042|
167
+ |house_sales|21613.0|17.0|https://www.openml.org/d/42731|https://www.openml.org/d/44066|
168
+ |seattlecrime6|52031.0|4.0|https://www.openml.org/d/42496|https://www.openml.org/d/45043|
169
+ |medical_charges|163065.0|5.0|https://www.openml.org/d/42720|https://www.openml.org/d/45048|
170
+ |particulate-matter-ukair-2017|394299.0|6.0|https://www.openml.org/d/42207|https://www.openml.org/d/44068|
171
+ |SGEMM_GPU_kernel_performance|241600.0|9.0|https://www.openml.org/d/43144|https://www.openml.org/d/44069|
172
 
173
 
174
  ### Dataset Curators
 
183
 
184
  Léo Grinsztajn, Edouard Oyallon, Gaël Varoquaux. Why do tree-based models still outperform deep
185
  learning on typical tabular data?. NeurIPS 2022 Datasets and Benchmarks Track, Nov 2022, New
186
+ Orleans, United States. ffhal-03723551v2f
clf_cat/KDDCup09_upselling.csv DELETED
The diff for this file is too large to render. See raw diff
 
clf_cat/albert.csv ADDED
The diff for this file is too large to render. See raw diff
 
clf_cat/compas-two-years.csv ADDED
The diff for this file is too large to render. See raw diff
 
clf_cat/compass.csv DELETED
The diff for this file is too large to render. See raw diff
 
clf_cat/covertype.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:246f5d269dde5ead8fa1462d73583de0459d3f6093ad61c9e38f6c156606364f
3
- size 62775986
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:716928e78f3af7947d20867ab7159f8f19e9a65d10bb681b676e6f2719fb9ece
3
+ size 59921335
clf_cat/default-of-credit-card-clients.csv ADDED
The diff for this file is too large to render. See raw diff
 
clf_cat/electricity.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_cat/eye_movements.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_cat/rl.csv DELETED
The diff for this file is too large to render. See raw diff
 
clf_cat/road-safety.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fc566ce4e8244946d0721dc22cc98308aaf058331c6844017a6afb393bcca7fb
3
- size 14561149
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c10d2cc6d04f8586a57a68ec5093c9420776ce08536a7fbe63c36fea15e586be
3
+ size 12995826
clf_num/Bioresponse.csv ADDED
The diff for this file is too large to render. See raw diff
 
clf_num/Diabetes130US.csv ADDED
The diff for this file is too large to render. See raw diff
 
clf_num/Higgs.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:97e7cfe445c3a0ae3809b41eff04065c588e75968bd20a0485ae5d00697efd08
3
- size 440107695
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc6eeaa6806da38dd272537c2d1d3646e1615c7e234f4eeea8920a4a7b719f24
3
+ size 433637684
clf_num/MagicTelescope.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/MiniBooNE.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c6ec9fdaad755c59cf276baf049599bf974b002848f131f5e7323005b132495b
3
- size 41225150
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f080afdc4b5fc29e29156bef893546a430e7ee8e89298416a46ccc92577279e
3
+ size 40798271
clf_num/bank-marketing.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/california.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/covertype.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:11d5086b4bba5013ab459ff876307efdb6b9db80c70ae9e48b0da667603ab0a7
3
- size 109908675
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d5a30a42f555ef8789f6511a2be15297335364336529bba28fb31d281f9068a
3
+ size 106053570
clf_num/credit.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/default-of-credit-card-clients.csv ADDED
The diff for this file is too large to render. See raw diff
 
clf_num/electricity.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/eye_movements.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/heloc.csv ADDED
The diff for this file is too large to render. See raw diff
 
clf_num/house_16H.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/jannis.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1c6e8a98cc3e6676ce28a4f2ef3a810a24467d9c0940e1ca026a67d0b1ea8a99
3
- size 26718488
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ec0a6d60f4fc6025eb8b1ad678f196d8c662982486f8a7dd91ee6a129e28b80
3
+ size 26384117
clf_num/kdd_ipums_la_97-small.csv DELETED
The diff for this file is too large to render. See raw diff
 
clf_num/phoneme.csv DELETED
The diff for this file is too large to render. See raw diff
 
clf_num/pol.csv CHANGED
The diff for this file is too large to render. See raw diff
 
clf_num/wine.csv DELETED
The diff for this file is too large to render. See raw diff
 
git ADDED
File without changes
reg_cat/{OnlineNewsPopularity.csv → Airlines_DepDelay_1M.csv} RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8d8985ded69f213ed2ff94d7c6683100c374552ce5251d40b7fb733e63a2004
3
- size 18853702
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fada6a98b181d14dd8e4dcb65587cc799ef5237f2adfa0f779e61489b236791e
3
+ size 41023434
reg_num/isolet.csv → reg_cat/Allstate_Claims_Severity.csv RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c35d63ff318c0660c267c1c7e0e1759581362793aeaca6fb2397f9a1f4910ab5
3
- size 32568247
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e47ef327aad9f06e5d7e03acc25d287edbb6eea3c1471f92a8ddce420b1287f8
3
+ size 76044510
reg_cat/Bike_Sharing_Demand.csv CHANGED
The diff for this file is too large to render. See raw diff
 
reg_cat/Brazilian_houses.csv CHANGED
The diff for this file is too large to render. See raw diff
 
reg_cat/Mercedes_Benz_Greener_Manufacturing.csv CHANGED
The diff for this file is too large to render. See raw diff
 
reg_cat/SGEMM_GPU_kernel_performance.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c2e98b1ffe16175bde3c54fb25225ff336006968f8f8f666ea31aaab73c0f734
3
- size 13482781
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb764f64d2e95a8856bc8de0b3dacf95d83a617f82fe72340698817735c79474
3
+ size 11902690
reg_cat/abalone.csv ADDED
The diff for this file is too large to render. See raw diff
 
reg_cat/analcatdata_supreme.csv CHANGED
The diff for this file is too large to render. See raw diff
 
reg_num/year.csv → reg_cat/delays_zurich_transport.csv RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6bdd6b9e49833a2fd764910ca7a518186e65d7b9c4bfe9645b2ffb8b6a2b1817
3
- size 446919659
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45c0e5e72510428347f6ff905cd42072af697fefb193541c053966a4637dd930
3
+ size 310520329
reg_cat/diamonds.csv CHANGED
The diff for this file is too large to render. See raw diff
 
reg_cat/house_sales.csv CHANGED
The diff for this file is too large to render. See raw diff
 
reg_cat/{black_friday.csv → medical_charges.csv} RENAMED
The diff for this file is too large to render. See raw diff
 
reg_cat/nyc-taxi-green-dec-2016.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c2bd9e5235795645bffb297be7106afbd403512539b06af0246c34af61dc8fc0
3
- size 37924077
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1532e83c397e21795bae42953eeed34a1ddd194030cdcf57f82a69baff9e45a5
3
+ size 33962341
reg_cat/particulate-matter-ukair-2017.csv CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:60efa8c8e822d005dbda9642249edc2afce6a82bd2b65a55d5917c6a23d09bb1
3
- size 17155359
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff67472d80bde9b82c1a1d23c962a83eda29a12c17422d40265456b7043f0d08
3
+ size 14506375
reg_cat/seattlecrime6.csv ADDED
The diff for this file is too large to render. See raw diff
 
reg_cat/topo_2_1.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a7dbd3d090926e8c4316826087d702c44fdacd9bb6b102e3c1f1a2b1e58f156
3
+ size 18340394
reg_cat/visualizing_soil.csv CHANGED
The diff for this file is too large to render. See raw diff
 
reg_cat/yprop_4_1.csv DELETED
The diff for this file is too large to render. See raw diff