We applied consensus clustering [ 36], a resampling-based method to estimate classification stability and classification accuracy (See Methods for details).
The weighted bags of visual words representation is coupled with a discriminative method to perform classification.
Next, we have applied the proposed method to material classification of a printed circuit board with various tiny elements.
VLAD-FV can be used instead as an off-the-shelf encoding method to improve classification accuracy when Fisher Vector or VLAD is going to be used.
For corporate credit, Altman 1968 and Baever 1966 applied a uni-variate model and a multiple DA method to the classification and prediction of business creditworthiness.
In addition, we investigate how many frames are necessary for our method to perform classification in the scenario of online action recognition.
We apply the proposed method to clinical classification based on gene-expression microarrays, demonstrating that the proposed method can outperform existing nearest centroid classifiers.
We apply our method to the classification of breast cancer metastasis, and demonstrate that it can achieve higher accuracy compared to several previous pathway-based approaches.
We apply the proposed method to the classification of breast cancer metastasis, and show that it achieves higher accuracy and identifies more reproducible pathway markers compared to several existing pathway activity inference methods.
Tenfold cross validation is a widely used method to examine classification performance.
