The document compares the performance of various machine learning classifiers on four datasets when using different feature extraction methods. For each dataset, it lists the classification time in seconds for fine KNN, cubic SVM, quadratic SVM, weighted KNN, medium Gaussian SVM, cosine KNN, wide NN, medium KNN, cubic KNN and medium NN classifiers using features extracted with EfficientNetB0 and Inception-v3 with serial fusion optimizations. Bar charts also visualize the results.
The document compares the performance of various machine learning classifiers on four datasets when using different feature extraction methods. For each dataset, it lists the classification time in seconds for fine KNN, cubic SVM, quadratic SVM, weighted KNN, medium Gaussian SVM, cosine KNN, wide NN, medium KNN, cubic KNN and medium NN classifiers using features extracted with EfficientNetB0 and Inception-v3 with serial fusion optimizations. Bar charts also visualize the results.
Original Description:
In this document you can study different type of datasets
The document compares the performance of various machine learning classifiers on four datasets when using different feature extraction methods. For each dataset, it lists the classification time in seconds for fine KNN, cubic SVM, quadratic SVM, weighted KNN, medium Gaussian SVM, cosine KNN, wide NN, medium KNN, cubic KNN and medium NN classifiers using features extracted with EfficientNetB0 and Inception-v3 with serial fusion optimizations. Bar charts also visualize the results.
The document compares the performance of various machine learning classifiers on four datasets when using different feature extraction methods. For each dataset, it lists the classification time in seconds for fine KNN, cubic SVM, quadratic SVM, weighted KNN, medium Gaussian SVM, cosine KNN, wide NN, medium KNN, cubic KNN and medium NN classifiers using features extracted with EfficientNetB0 and Inception-v3 with serial fusion optimizations. Bar charts also visualize the results.
Download as XLSX, PDF, TXT or read online from Scribd
Download as xlsx, pdf, or txt
You are on page 1of 10
IXMAS Dataset
Classifiers Features Extractions with EfficientNetB0
Fine KNN 222.94 Cubic SVM 590.9 Quadratic SVM 520.55 Weighted KNN 224.79 Medium Gaussian SVM 848.74 Cosine KNN 220.51 Wide NN 70.305 Medium KNN 222.25 Cubic KNN 5105.4 Medium NN 54.764
14000
12000
10000
8000
6000
4000
2000
0 Fine KNN Cubic SVM Quadratic Weighted Medium Cosine KNN Wide NN Medium Cubic KNN SVM KNN Gaussian KNN SVM
Features Extractions with EfficientNetB0 Features Extractions with Inception-v3
Serial Fusion Optimizations
KTH Dataset Classifiers Features Extractions with EfficientNetB0 Fine KNN 2569.4 Cubic SVM 1141.6 Quadratic SVM 845.44 Weighted KNN 5068.7 Medium Gaussian SVM 2696.4 Cosine KNN 4253.6 Wide NN 5768.5 Medium KNN 3763.5 Cubic KNN 9839.3 Medium NN 5551.6
Chart Title 12000
10000
8000
6000
4000
2000
0 Fine KNN Cubic SVM Quadratic Weighted Medium Cosine KNN Wide NN Medium Cubic KNN SVM KNN Gaussian KNN SVM
Features Extractions with EfficientNetB0 Features Extractions with Inception-v3
Serial Fusion Optimizations
UCF_Sports Dataset Classifiers Features Extractions with EfficientNetB0 Fine KNN 282.75 Cubic SVM 460.12 Quadratic SVM 419.29 Weighted KNN 270.24 Medium Gaussian SVM 721.69 Cosine KNN 269.8 Wide NN 56.819 Medium KNN 265.79 Cubic KNN 6473.6 Medium NN 33.823
Chart Title 80000
70000
60000
50000
40000
30000
20000
10000
0 Fine KNN Cubic SVM Quadratic Weighted Medium Cosine KNN Wide NN Medium Cubic KNN SVM KNN Gaussian KNN SVM
Features Extractions with EfficientNetB0 Features Extractions with Inception-v3
Serial Fusion Optimizations
UCFYoutube Dataset Classifiers Features Extractions with EfficientNetB0 Fine KNN 2358.7 Cubic SVM 1267.6 Quadratic SVM 1231.3 Weighted KNN 2165.9 Medium Gaussian SVM 2103.4 Cosine KNN 3197.4 Wide NN 419.77 Medium KNN 1951 Cubic KNN 23263 Medium NN 51.184
Chart Title 70000
60000
50000
40000
30000
20000
10000
0 Fine KNN Cubic SVM Quadratic Weighted Medium Cosine KNN Wide NN Medium Cubic KNN SVM KNN Gaussian KNN SVM
Features Extractions with EfficientNetB0 Features Extractions with Inception-v3
Serial Fusion Optimizations
Weizmann Dataset Classifiers Features Extractions with EfficientNetB0 Fine KNN 87.575 Cubic SVM 130.19 Quadratic SVM 114.07 Weighted KNN 47.78 Medium Gaussian SVM 159.09 Cosine KNN 48.962 Wide NN 40.541 Medium KNN 48.566 Cubic KNN 985.56 Medium NN 31.043
Chart Title 2500 Chart Title 2500
2000
1500
1000
500
0 Fine KNN Cubic SVM Quadratic Weighted Medium Cosine KNN Wide NN Medium Cubic KNN SVM KNN Gaussian KNN SVM
Features Extractions with EfficientNetB0 Features Extractions with Inception-v3
