From: Comparing classification techniques for identification of grasped objects
Classification technique | Training time (ms) | Second classifier structure scenario 1 (universal use) | ||
---|---|---|---|---|
Model M1 | Model M2 | Test time (ms) | Accuracy | |
Bagging | 392.18 | 276.36 | 39.74 | 95.9 |
Decision tree (entropy) | 5.16 | 6.82 | 10.96 | 82.8 |
Decision tree (Gini) | 5.04 | 5.14 | 10.49 | 92.0 |
kNN | 0.84 | 0.78 | 19.72 | 94.7 |
Linear discriminant analysis | 5.62 | 6.96 | 11.35 | 67.1 |
SVM (linear SVC) | 146.66 | 202.21 | 12.47 | 69.1 |
Logistic regression | 16.21 | 59.15 | 11.58 | 73.6 |
Logistic regression CV | 314.46 | 1320.23 | 11.81 | 74.0 |
MLP | 2478.57 | 2886.71 | 12.07 | 90.5 |
Naive Bayes (Bernoulli) | 1.46 | 1.04 | 11.73 | 56.0 |
Naive Bayes (Gaussian) | 1.5 | 1.04 | 12.09 | 85.3 |
NearestCentroid | 1.02 | 0.6 | 11.49 | 64.3 |
Quadratic discriminant analysis | 1.52 | 1.4 | 11.87 | 82.5 |
Radius neighbors | 1.14 | 0.82 | 247.42 | 54.7 |
Random forest | 199.41 | 158.87 | 36.58 | 96.6 |
Ridge | 1.79 | 1.28 | 11.76 | 60.9 |
Ridge CV | 2.64 | 2.38 | 11.42 | 61.0 |
Label propagation | 56.54 | 57.31 | 46.22 | 96.7 |
Label spreading | 97.28 | 99.82 | 45.64 | 96.3 |