Yimin Fan / School of Information Engineering, Zhengzhou University
Lin Qi / School of Information Engineering, Zhengzhou University
Yun Tie / School of Information Engineering, Zhengzhou University
It is very important to classify some small-scale datasets accurately in biology. With the rapid advancement of classification models, such as support vector machine(SVM), Random Forest(RF), Deep Forest, Convolutional Neural Networks(CNNs), etc. However, for small-scale datasets, CNN always need massive datasets to train. Other methods usually can’t achieve better effects. Therefore, in this paper, a new forest model is proposed to solve the problems in small-scale datasets. It improves the classification performance through integrated learning method. The improved model distinguishes from the original model with two main aspects: That is, first, considering the fitting quality of each forest, the standard deviation of some most important features in each forest compose a new feature to be concatenated in the next cascade layer. Second, the sub-layer structure is adapted to the cascade layer to increase the training opportunities. Experiments on five datasets demonstrate that our model has better classification performance than other classification models in the small-scale datasets.