MLcomp version alpha Statistics: 1403 users, 461 datasets, 360 programs, 16577 runs (0 queued, 0 running), 0 workers

Run 13209

Creator	yorinasub17
Program	svmlight_multiclass
Dataset	Synthetic 75% Density, Large, Many Labels
Task type	MulticlassClassification
Created	28d16h ago
Download	Login required!

Status: Done!

Total Time	1m54s
Max Memory Usage	350M
Domain	MulticlassClassification
Learn time	1m17s
Train error	0.278
Predict train time	15s
Test error	0.437
Predict test time	7s

Log file

===== MAIN: learn based on training data =====
=== START program1: ./run learn ../dataset2/train
Using hyperparameter c = 0.1
Reading training examples... (5250 examples) done
Training set properties: 1000 features, 20 classes
Iter 1: .........*(NumConst=1, SV=1, CEps=100.0000, QPEps=0.0000)
Iter 2: *(NumConst=2, SV=2, CEps=40.9046, QPEps=0.0000)
Iter 3: *(NumConst=3, SV=2, CEps=22.9647, QPEps=0.0000)
Iter 4: .........*(NumConst=4, SV=2, CEps=186.2510, QPEps=20.7704)
Iter 5: *(NumConst=5, SV=3, CEps=27.5200, QPEps=0.0000)
Iter 6: *(NumConst=6, SV=4, CEps=34.6838, QPEps=0.0000)
Iter 7: .........*(NumConst=7, SV=6, CEps=81.6540, QPEps=38.8172)
Iter 8: *(NumConst=8, SV=6, CEps=59.5027, QPEps=9.7928)
Iter 9: *(NumConst=9, SV=6, CEps=77.9684, QPEps=23.5573)
Iter 10: *(NumConst=10, SV=6, CEps=51.0789, QPEps=0.0122)
Iter 11: *(NumConst=11, SV=6, CEps=55.9685, QPEps=0.2182)
Iter 12: *(NumConst=12, SV=7, CEps=11.9127, QPEps=0.0000)
Iter 13: *(NumConst=13, SV=8, CEps=15.3789, QPEps=0.0159)
Iter 14: *(NumConst=14, SV=8, CEps=12.4697, QPEps=0.0001)
Iter 15: *(NumConst=15, SV=7, CEps=20.5442, QPEps=0.1700)
Iter 16: .........*(NumConst=16, SV=10, CEps=57.9060, QPEps=8.2994)
Iter 17: *(NumConst=17, SV=11, CEps=30.3090, QPEps=12.0529)
Iter 18: *(NumConst=18, SV=8, CEps=10.7426, QPEps=0.7991)
Iter 19: *(NumConst=19, SV=8, CEps=28.0114, QPEps=0.0030)
Iter 20: *(NumConst=20, SV=7, CEps=21.1828, QPEps=0.0000)
Iter 21: *(NumConst=21, SV=8, CEps=9.1229, QPEps=0.0004)
Iter 22: *(NumConst=22, SV=7, CEps=9.0141, QPEps=0.0054)
Iter 23: *(NumConst=23, SV=9, CEps=11.5717, QPEps=1.9927)
Iter 24: *(NumConst=24, SV=8, CEps=9.1036, QPEps=2.4602)
Iter 25: .........*(NumConst=25, SV=9, CEps=58.1699, QPEps=3.8183)
Iter 26: *(NumConst=26, SV=10, CEps=9.7104, QPEps=0.0000)
Iter 27: *(NumConst=27, SV=11, CEps=6.2088, QPEps=0.0000)
Iter 28: .........*(NumConst=28, SV=12, CEps=29.0434, QPEps=0.0000)
Iter 29: *(NumConst=29, SV=13, CEps=10.0796, QPEps=3.8283)
Iter 30: *(NumConst=30, SV=13, CEps=10.5002, QPEps=3.4269)
Iter 31: *(NumConst=31, SV=13, CEps=4.3151, QPEps=0.1597)
Iter 32: *(NumConst=32, SV=15, CEps=3.5598, QPEps=0.9931)
Iter 33: .........*(NumConst=33, SV=16, CEps=18.4807, QPEps=2.6557)
Iter 34: *(NumConst=34, SV=17, CEps=6.0321, QPEps=2.9372)
Iter 35: *(NumConst=35, SV=14, CEps=6.5009, QPEps=0.8969)
Iter 36: *(NumConst=36, SV=14, CEps=2.8306, QPEps=0.1710)
Iter 37: *(NumConst=37, SV=14, CEps=2.4596, QPEps=0.4464)
Iter 38: .........*(NumConst=38, SV=15, CEps=8.5592, QPEps=0.1072)
Iter 39: *(NumConst=39, SV=16, CEps=3.5190, QPEps=0.9043)
Iter 40: *(NumConst=40, SV=16, CEps=3.9213, QPEps=0.7249)
Iter 41: *(NumConst=41, SV=16, CEps=2.0567, QPEps=0.6639)
Iter 42: *(NumConst=42, SV=15, CEps=1.0487, QPEps=0.3444)
Iter 43: *(NumConst=43, SV=13, CEps=1.8852, QPEps=0.0000)
Iter 44: *(NumConst=44, SV=15, CEps=1.1730, QPEps=0.4430)
Iter 45: *(NumConst=45, SV=16, CEps=1.3385, QPEps=0.3497)
Iter 46: .........*(NumConst=46, SV=17, CEps=2.9364, QPEps=0.3104)
Iter 47: *(NumConst=47, SV=16, CEps=1.5869, QPEps=0.3975)
Iter 48: *(NumConst=48, SV=16, CEps=1.6478, QPEps=0.6159)
Iter 49: *(NumConst=49, SV=16, CEps=1.3025, QPEps=0.2576)
Iter 50: *(NumConst=50, SV=16, CEps=1.3304, QPEps=0.4720)
Iter 51: *(NumConst=51, SV=16, CEps=0.9137, QPEps=0.3785)
Iter 52: *(NumConst=52, SV=17, CEps=0.8843, QPEps=0.4175)
Iter 53: *(NumConst=53, SV=17, CEps=1.2755, QPEps=0.3919)
Iter 54: *(NumConst=54, SV=18, CEps=1.2653, QPEps=0.4005)
Iter 55: *(NumConst=55, SV=19, CEps=0.6300, QPEps=0.2663)
Iter 56: *(NumConst=56, SV=20, CEps=0.8706, QPEps=0.3142)
Iter 57: *(NumConst=56, SV=20, CEps=0.7228, QPEps=0.2302)
Iter 58: *(NumConst=56, SV=21, CEps=0.5242, QPEps=0.1163)
Iter 59: *(NumConst=57, SV=22, CEps=0.5431, QPEps=0.2595)
Iter 60: *(NumConst=56, SV=21, CEps=0.4531, QPEps=0.2033)
Iter 61: *(NumConst=57, SV=21, CEps=0.5279, QPEps=0.1296)
Iter 62: *(NumConst=58, SV=22, CEps=0.3161, QPEps=0.1265)
Iter 63: *(NumConst=58, SV=21, CEps=0.5318, QPEps=0.0719)
Iter 64: .........*(NumConst=59, SV=22, CEps=2.1447, QPEps=0.3309)
Iter 65: *(NumConst=60, SV=22, CEps=1.4989, QPEps=0.5567)
Iter 66: *(NumConst=61, SV=21, CEps=1.5929, QPEps=0.5261)
Iter 67: *(NumConst=57, SV=22, CEps=1.1043, QPEps=0.4621)
Iter 68: *(NumConst=57, SV=25, CEps=1.0676, QPEps=0.5084)
Iter 69: *(NumConst=57, SV=22, CEps=0.6152, QPEps=0.1996)
Iter 70: *(NumConst=58, SV=22, CEps=0.8557, QPEps=0.2801)
Iter 71: *(NumConst=58, SV=22, CEps=0.6234, QPEps=0.2871)
Iter 72: *(NumConst=58, SV=21, CEps=0.4030, QPEps=0.1786)
Iter 73: *(NumConst=57, SV=23, CEps=0.7965, QPEps=0.1791)
Iter 74: *(NumConst=58, SV=21, CEps=0.5135, QPEps=0.1957)
Iter 75: *(NumConst=59, SV=19, CEps=0.5492, QPEps=0.1083)
Iter 76: *(NumConst=60, SV=23, CEps=0.3908, QPEps=0.1475)
Iter 77: *(NumConst=61, SV=23, CEps=0.6247, QPEps=0.1877)
Iter 78: *(NumConst=62, SV=26, CEps=0.5586, QPEps=0.1326)
Iter 79: *(NumConst=62, SV=23, CEps=0.2929, QPEps=0.0747)
Iter 80: *(NumConst=63, SV=22, CEps=0.2920, QPEps=0.0913)
Iter 81: *(NumConst=64, SV=26, CEps=0.3081, QPEps=0.1338)
Iter 82: .........*(NumConst=65, SV=27, CEps=0.9747, QPEps=0.3318)
Iter 83: *(NumConst=66, SV=29, CEps=0.7842, QPEps=0.3828)
Iter 84: *(NumConst=63, SV=32, CEps=0.7221, QPEps=0.3233)
Iter 85: *(NumConst=63, SV=30, CEps=0.6213, QPEps=0.2987)
Iter 86: *(NumConst=64, SV=28, CEps=0.5832, QPEps=0.2908)
Iter 87: *(NumConst=65, SV=24, CEps=0.4788, QPEps=0.2272)
Iter 88: *(NumConst=66, SV=24, CEps=0.6870, QPEps=0.1839)
Iter 89: *(NumConst=66, SV=25, CEps=0.5294, QPEps=0.2143)
Iter 90: *(NumConst=67, SV=27, CEps=0.4615, QPEps=0.1509)
Iter 91: *(NumConst=67, SV=27, CEps=0.4724, QPEps=0.2085)
Iter 92: *(NumConst=65, SV=25, CEps=0.3937, QPEps=0.1577)
Iter 93: *(NumConst=66, SV=27, CEps=0.3897, QPEps=0.1767)
Iter 94: *(NumConst=66, SV=26, CEps=0.3091, QPEps=0.1045)
Iter 95: *(NumConst=67, SV=27, CEps=0.2688, QPEps=0.1304)
Iter 96: *(NumConst=67, SV=27, CEps=0.3235, QPEps=0.1330)
Iter 97: *(NumConst=68, SV=26, CEps=0.2391, QPEps=0.1180)
Iter 98: *(NumConst=68, SV=24, CEps=0.2615, QPEps=0.0595)
Iter 99: *(NumConst=68, SV=25, CEps=0.3078, QPEps=0.1053)
Iter 100: *(NumConst=69, SV=30, CEps=0.2385, QPEps=0.0971)
Iter 101: *(NumConst=70, SV=28, CEps=0.2023, QPEps=0.1003)
Iter 102: *(NumConst=71, SV=29, CEps=0.1756, QPEps=0.0805)
Iter 103: *(NumConst=72, SV=30, CEps=0.1705, QPEps=0.0751)
Iter 104: *(NumConst=72, SV=30, CEps=0.2458, QPEps=0.0716)
Iter 105: *(NumConst=73, SV=30, CEps=0.1548, QPEps=0.0772)
Iter 106: *(NumConst=74, SV=30, CEps=0.2502, QPEps=0.0576)
Iter 107: *(NumConst=75, SV=31, CEps=0.1893, QPEps=0.0748)
Iter 108: *(NumConst=76, SV=30, CEps=0.1335, QPEps=0.0539)
Iter 109: *(NumConst=77, SV=30, CEps=0.1293, QPEps=0.0534)
Iter 110: *(NumConst=78, SV=30, CEps=0.1220, QPEps=0.0507)
Iter 111: *(NumConst=79, SV=27, CEps=0.1065, QPEps=0.0522)
Iter 112: *(NumConst=80, SV=30, CEps=0.1408, QPEps=0.0460)
Iter 113: *(NumConst=81, SV=30, CEps=0.1065, QPEps=0.0529)
Iter 114: .........*(NumConst=82, SV=31, CEps=0.3359, QPEps=0.1498)
Iter 115: *(NumConst=82, SV=33, CEps=0.2317, QPEps=0.1141)
Iter 116: *(NumConst=82, SV=37, CEps=0.2936, QPEps=0.1149)
Iter 117: *(NumConst=82, SV=39, CEps=0.1961, QPEps=0.0959)
Iter 118: *(NumConst=82, SV=33, CEps=0.1906, QPEps=0.0863)
Iter 119: *(NumConst=83, SV=36, CEps=0.2058, QPEps=0.0925)
Iter 120: *(NumConst=84, SV=36, CEps=0.1987, QPEps=0.0899)
Iter 121: *(NumConst=85, SV=36, CEps=0.1688, QPEps=0.0680)
Iter 122: *(NumConst=86, SV=36, CEps=0.1484, QPEps=0.0609)
Iter 123: *(NumConst=85, SV=38, CEps=0.1098, QPEps=0.0540)
Iter 124: *(NumConst=86, SV=38, CEps=0.1006, QPEps=0.0471)
Iter 125: *(NumConst=86, SV=38, CEps=0.1146, QPEps=0.0469)
Iter 126: .........*(NumConst=86, SV=37, CEps=0.1411, QPEps=0.0651)
Iter 127: *(NumConst=87, SV=36, CEps=0.1033, QPEps=0.0502)
Iter 128: *(NumConst=87, SV=36, CEps=0.1021, QPEps=0.0505)
Iter 129: *(NumConst=87, SV=38, CEps=0.1112, QPEps=0.0441)
Iter 130: .........(NumConst=87, SV=38, CEps=0.0876, QPEps=0.0441)
Final epsilon on KKT-Conditions: 0.08759
Upper bound on duality gap: 0.00803
Dual objective value: dval=8.28822
Primal objective value: pval=8.29625
Total number of constraints in final working set: 87 (of 129)
Number of iterations: 130
Number of calls to 'find_most_violated_constraint': 73500
Number of SV: 38 
Norm of weight vector: |w|=0.59587
Value of slack variable (on working set): xi=81.12875
Value of slack variable (global): xi=81.18717
Norm of longest difference vector: ||Psi(x,y)-Psi(x,ybar)||=221.13547
Runtime in cpu-seconds: 23.44
Final number of constraints in cache: 26108
Compacting linear model...done
Writing learned model...done
=== END program1: ./run learn ../dataset2/train --- OK [77s]

===== MAIN: predict/evaluate on train data =====
=== START program3: ./run stripLabels ../dataset2/train ../program0/evalTrain.in
=== END program3: ./run stripLabels ../dataset2/train ../program0/evalTrain.in --- OK [6s]
=== START program1: ./run predict ../program0/evalTrain.in ../program0/evalTrain.out
Reading model...done.
Reading test examples... (5250 examples) done.
Classifying test examples...done
Runtime (without IO) in cpu-seconds: 0.32
Average loss on test set: 99.2190
Zero/one-error on test set: 99.22% (41 correct, 5209 incorrect, 5250 total)
=== END program1: ./run predict ../program0/evalTrain.in ../program0/evalTrain.out --- OK [15s]
=== START program4: ./run evaluate ../dataset2/train ../program0/evalTrain.out
=== END program4: ./run evaluate ../dataset2/train ../program0/evalTrain.out --- OK [6s]

===== MAIN: predict/evaluate on test data =====
=== START program3: ./run stripLabels ../dataset2/test ../program0/evalTest.in
=== END program3: ./run stripLabels ../dataset2/test ../program0/evalTest.in --- OK [2s]
=== START program1: ./run predict ../program0/evalTest.in ../program0/evalTest.out
Reading model...done.
Reading test examples... (2250 examples) done.
Classifying test examples...done
Runtime (without IO) in cpu-seconds: 0.13
Average loss on test set: 99.5556
Zero/one-error on test set: 99.56% (10 correct, 2240 incorrect, 2250 total)
=== END program1: ./run predict ../program0/evalTest.in ../program0/evalTest.out --- OK [7s]
=== START program4: ./run evaluate ../dataset2/test ../program0/evalTest.out
=== END program4: ./run evaluate ../dataset2/test ../program0/evalTest.out --- OK [2s]


real	1m55.551s
user	0m41.779s
sys	0m1.348s

Run specification

• supervised-learning: Main entry for supervised learning for training and testing a program on a dataset.
  • (learner:Program) svmlight_multiclass: SVMlight for multiclass classification (http://svmlight.joachims.org/svm_multiclass.html)
  • (dataset:Dataset) Synthetic 75% Density, Large, Many Labels: A synthetically generated data set Attributes: =label(i) = argmax_j w(j)'*x(i) for randomly generated weight vectors. =weight vectors elements are independently sampled from Normal distribution. =density is what percentage of weight vector's elements were not set to 0 =x(i) normally distributed according to multivariate Gaussian, random parameters
  • (stripper:Program[Strip]) multiclass-utils: Validates and inspects a dataset in MulticlassClassification format.
  • (evaluator:Program[Evaluate]) classification-evaluator: Evaluates predictions of classification datasets (discrete outputs).

Results

• doTest:
  • evaluate:
    • errorRate: 0.436888888888889
    • numErrors: 983
    • numExamples: 2250
    • success: true
    • time: 2
  • predict:
    • success: true
    • time: 7
  • strip:
    • success: true
    • time: 2
• doTrain:
  • evaluate:
    • errorRate: 0.278095238095238
    • numErrors: 1460
    • numExamples: 5250
    • success: true
    • time: 6
  • predict:
    • success: true
    • time: 15
  • strip:
    • success: true
    • time: 6
• exitCode: 0
• learn:
  • success: true
  • time: 77
• success: true

Comments: