《Table 1 Prediction results of several methods w ith ACC transformation》

  提示：宽带有限、当前游客访问压缩模式

本系列图表出处文件名：随高清版一同展现

《Prediction of Protein-Protein Interactions by a Novel Model Based on Domain Information》

1. 获取 高清版本忘记账户？点击这里登录

1. 下载图表忘记账户？点击这里登录

In order to validate the proposed model，this paper conducts a series of comparative tests of different algorithms w ith ACC and Pse AAC on 5 datasets generated by section 3.1and the averages of the five results are taken as the final results.For TM LM，four typical machine learning algorithms，Naive Bayes，Ada Boost M1，Random Tree and Support Vector M achine（SVM）are employed.For BM ILM，3 traditional famous algorithms，BCk NN，M i Ck NN and Bagging_C_k NN are chosen.For IM LM，6 representative algorithms are chosen，four of them are classical M IL algorithms，M ILD_B，M ILIS，M ILES and M ILD_I，other tw o algorithms are M il Ca w hich has been proved to be an efficient method and M il Ca A presented in this paper.In addition，ACC is also used in the step of feature encoding to compare w ith Pse ACC adopted by our model.The prediction results of above methods are given in Tables 1 and 2，respectively.The best performances are highlighted in bold.And the corresponding curve graphs are rendered in Figs.1 and2，to give a more clear and intuitive presentation of the results.