The overall process flow of our CP is depicted in Fig. 1, with the proposed CP-EGBM classification model in red. The following sub-sections provide the details of the model.
Data preprocessing and feature selection
Let the dataset consist of N examples of M-dimension feature vectors \(\left\{{x}_{n,m}, 1\le n\le N \mathrm{and} 1\le m\le M\right\}\) and target label \(\left\{y, 1\le y\le C\right\}\) where C is the number of classes. Each feature in the dataset is normalized in the range [0, 1] as per Eq. (1) to improve…
Continue Reading
News Source: www.nature.com