Skip to content

Commit 2432ead

Browse files
AlexsandrussAlexsandruss
committed
Additional configs
1 parent fd6578e commit 2432ead

File tree

5 files changed

+736
-34
lines changed

5 files changed

+736
-34
lines changed

‎.gitignore

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -10,4 +10,5 @@ __work*
1010

1111
# Datasets
1212
dataset
13+
data
1314
*.csv

‎configs/cuml_config.json

Lines changed: 173 additions & 32 deletions
Original file line numberDiff line numberDiff line change
@@ -99,6 +99,44 @@
9999
"svd-solver": ["full"],
100100
"n-components": [10]
101101
},
102+
{
103+
"algorithm": "df_clsf",
104+
"dtype": ["float32"],
105+
"dataset": [
106+
{
107+
"source": "csv",
108+
"name": "higgs1m",
109+
"training":
110+
{
111+
"x": "data/higgs1m_x_train.csv",
112+
"y": "data/higgs1m_y_train.csv"
113+
},
114+
"testing":
115+
{
116+
"x": "data/higgs1m_x_test.csv",
117+
"y": "data/higgs1m_y_test.csv"
118+
}
119+
},
120+
{
121+
"source": "csv",
122+
"name": "airline-ohe",
123+
"training":
124+
{
125+
"x": "data/airline-ohe_x_train.csv",
126+
"y": "data/airline-ohe_y_train.csv"
127+
},
128+
"testing":
129+
{
130+
"x": "data/airline-ohe_x_test.csv",
131+
"y": "data/airline-ohe_y_test.csv"
132+
}
133+
}
134+
],
135+
"num-trees": [50],
136+
"max-depth": [16],
137+
"max-leaf-nodes": [131072],
138+
"max-features": [0.2]
139+
},
102140
{
103141
"algorithm": "ridge",
104142
"dataset": [
@@ -185,6 +223,90 @@
185223
"maxiter": [100],
186224
"tol": [0]
187225
},
226+
{
227+
"algorithm": "svm",
228+
"dataset": [
229+
{
230+
"source": "csv",
231+
"name": "ijcnn",
232+
"training":
233+
{
234+
"x": "data/ijcnn_x_train.csv",
235+
"y": "data/ijcnn_y_train.csv"
236+
},
237+
"testing":
238+
{
239+
"x": "data/ijcnn_x_test.csv",
240+
"y": "data/ijcnn_y_test.csv"
241+
}
242+
}
243+
],
244+
"C": [1000.0],
245+
"kernel": ["linear"]
246+
},
247+
{
248+
"algorithm": "svm",
249+
"dataset": [
250+
{
251+
"source": "csv",
252+
"name": "a9a",
253+
"training":
254+
{
255+
"x": "data/a9a_x_train.csv",
256+
"y": "data/a9a_y_train.csv"
257+
},
258+
"testing":
259+
{
260+
"x": "data/a9a_x_test.csv",
261+
"y": "data/a9a_y_test.csv"
262+
}
263+
}
264+
],
265+
"C": [500.0],
266+
"kernel": ["rbf"]
267+
},
268+
{
269+
"algorithm": "svm",
270+
"dataset": [
271+
{
272+
"source": "csv",
273+
"name": "gisette",
274+
"training":
275+
{
276+
"x": "data/gisette_x_train.csv",
277+
"y": "data/gisette_y_train.csv"
278+
},
279+
"testing":
280+
{
281+
"x": "data/gisette_x_test.csv",
282+
"y": "data/gisette_y_test.csv"
283+
}
284+
}
285+
],
286+
"C": [1.5e-3],
287+
"kernel": ["linear"]
288+
},
289+
{
290+
"algorithm": "svm",
291+
"dataset": [
292+
{
293+
"source": "csv",
294+
"name": "klaverjas",
295+
"training":
296+
{
297+
"x": "data/klaverjas_x_train.csv",
298+
"y": "data/klaverjas_y_train.csv"
299+
},
300+
"testing":
301+
{
302+
"x": "data/klaverjas_x_test.csv",
303+
"y": "data/klaverjas_y_test.csv"
304+
}
305+
}
306+
],
307+
"C": [1.0],
308+
"kernel": ["rbf"]
309+
},
188310
{
189311
"algorithm": "dbscan",
190312
"dataset": [
@@ -218,73 +340,92 @@
218340
]
219341
},
220342
{
221-
"algorithm": "knn_clsf",
222-
"dtype": ["float32"],
343+
"algorithm": "train_test_split",
223344
"dataset": [
224345
{
225346
"source": "synthetic",
226347
"type": "classification",
227348
"n_classes": 2,
228-
"n_features": 3,
349+
"n_features": 20,
229350
"training": {
230-
"n_samples": 100000
351+
"n_samples": 5000000
231352
}
232353
},
233354
{
234355
"source": "synthetic",
235356
"type": "classification",
236357
"n_classes": 2,
237-
"n_features": 10,
358+
"n_features": 100,
238359
"training": {
239-
"n_samples": 100000
360+
"n_samples": 1000000
240361
}
241362
},
242363
{
243364
"source": "synthetic",
244365
"type": "classification",
245366
"n_classes": 2,
246-
"n_features": 50,
367+
"n_features": 10000,
247368
"training": {
248-
"n_samples": 20000
369+
"n_samples": 10000
249370
}
250371
}
251372
],
252-
"weights": ["uniform"],
253-
"method": ["kd_tree", "brute"]
373+
"train-size": [0.75],
374+
"test-size": [0.25]
254375
},
255376
{
256377
"algorithm": "train_test_split",
257378
"dataset": [
258379
{
259-
"source": "synthetic",
260-
"type": "classification",
261-
"n_classes": 2,
262-
"n_features": 20,
263-
"training": {
264-
"n_samples": 5000000
380+
"source": "csv",
381+
"name": "census",
382+
"training":
383+
{
384+
"x": "data/census_x.csv",
385+
"y": "data/census_y.csv"
265386
}
266-
},
387+
}
388+
],
389+
"train-size": [0.9],
390+
"test-size": [0.1]
391+
},
392+
{
393+
"algorithm": "lasso",
394+
"dataset": [
267395
{
268-
"source": "synthetic",
269-
"type": "classification",
270-
"n_classes": 2,
271-
"n_features": 100,
272-
"training": {
273-
"n_samples": 1000000
396+
"source": "csv",
397+
"name": "mortgage",
398+
"training":
399+
{
400+
"x": "data/mortgage_x.csv",
401+
"y": "data/mortgage_y.csv"
274402
}
275-
},
403+
}
404+
],
405+
"alpha": [1.0],
406+
"tol": [1e-4]
407+
},
408+
{
409+
"algorithm": "elasticnet",
410+
"dataset": [
276411
{
277-
"source": "synthetic",
278-
"type": "classification",
279-
"n_classes": 2,
280-
"n_features": 10000,
281-
"training": {
282-
"n_samples": 10000
412+
"source": "csv",
413+
"name": "year_prediction_msd",
414+
"training":
415+
{
416+
"x": "data/year_prediction_msd_x_train.csv",
417+
"y": "data/year_prediction_msd_y_train.csv"
418+
},
419+
"testing":
420+
{
421+
"x": "data/year_prediction_msd_x_test.csv",
422+
"y": "data/year_prediction_msd_y_test.csv"
283423
}
284424
}
285425
],
286-
"train-size": [0.75],
287-
"test-size": [0.25]
426+
"alpha": [2.0],
427+
"l1_ratio": [0.5],
428+
"tol": [1e-4]
288429
}
289430
]
290431
}

0 commit comments

Comments
 (0)