<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Feature ranking can not only help the data analysis system improve efficiency, but also reduce the interference of redundant features and irrelevant features to the results. At present, feature ranking of massive data is an important and difficult problem. In order to solve the above problems, this paper proposes a feature importance ranking algorithm based on XGBoost and MIC model by analyzing the existing algorithm models. Firstly, XGBoost model and MIC model are established respectively; Then, the results of the above two models are weighted and combined by the error reciprocal method. XGBoost model has the advantages of high efficiency, flexibility and portability, while MIC model has universality and easy parameter adjustment. The resulting XGBoost MIC combination model has both advantages; Finally, the first mock exam is used as a sample set of data for anticancer drug candidates. After preprocessing the data set, the XGBoost-MIC combination model is used to analyze the case. At the same time, the calculation results of a single model are calculated, and the model is optimized by adjusting the parameters of the model. The results show that the error of the first mock exam is obviously lower than that of the single calculation model, and the accuracy of the XGBoost-MIC is 0.75, which is 0.02 higher than that of the single model.</p>
</abstract>

Feature ranking can not only help the data analysis system improve efficiency, but also reduce the interference of redundant features and irrelevant features to the results. At present, feature ranking of massive data is an important and difficult problem. In order to solve the above problems, this paper proposes a feature importance ranking algorithm based on XGBoost and MIC model by analyzing the existing algorithm models. Firstly, XGBoost model and MIC model are established respectively; Then, the results of the above two models are weighted and combined by the error reciprocal method. XGBoost model has the advantages of high efficiency, flexibility and portability, while MIC model has universality and easy parameter adjustment. The resulting XGBoost MIC combination model has both advantages; Finally, the first mock exam is used as a sample set of data for anticancer drug candidates. After preprocessing the data set, the XGBoost-MIC combination model is used to analyze the case. At the same time, the calculation results of a single model are calculated, and the model is optimized by adjusting the parameters of the model. The results show that the error of the first mock exam is obviously lower than that of the single calculation model, and the accuracy of the XGBoost-MIC is 0.75, which is 0.02 higher than that of the single model.

Feature Sorting Algorithm Based on XGBoost and MIC Combination Model

School of Computer Science and Engineering

Jinan University-University of Birmingham Joint Institute

International Journal of Advanced Network, Monitoring and Controls

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

{"article-title":"Feature Sorting Algorithm Based on XGBoost and MIC Combination Model"}

Feature ranking can not only help the data analysis system improve efficiency, but also reduce the interference of redundant features and irrelevant features...

Method	R²	RMSE
composite pattern	0.759	8.705
XGBoost model	0.723	9.453
MIC model	0.689	8.994

Feature Sorting Algorithm Based on XGBoost and MIC Combination Model

Online veröffentlicht: 21. Mai 2023

Seitenbereich: 79 - 88

DOI: https://doi.org/10.21307/ijanmc-2021-037

Schlüsselwörter
Feature Sorting, MIC Arithmetic, Xgboost Arithmetic, Combination Model

© 2021 Gao Xiang et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Figure 1

Figure 2

Figure 3

Figure 4

Comparsion of R2 and RMSE values between XGBoost-MIC conbined model and single model.

Feature Sorting Algorithm Based on XGBoost and MIC Combination Model

Online veröffentlicht: 21. Mai 2023

Seitenbereich: 79 - 88

DOI: https://doi.org/10.21307/ijanmc-2021-037

SchlüsselwörterFeature Sorting, MIC Arithmetic, Xgboost Arithmetic, Combination Model

© 2021 Gao Xiang et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Figure 1

Figure 2

Figure 3

Figure 4

Comparsion of R2 and RMSE values between XGBoost-MIC conbined model and single model.

Schlüsselwörter
Feature Sorting, MIC Arithmetic, Xgboost Arithmetic, Combination Model