建構多階段混合式分類架構於肝癌患者復發預測之研究

CONSTRUCT A MULTI-STAGE HYBRID CLASSIFICATION FRAMEWORK TO PREDICT THE RECURRENCE OF LIVER CANCER PATIENTS

李岳樺(Yueh-Hua Lee) ； 沈湘莉(Hsiang-Li Shen) ； 呂奇傑(Chi-Jie Lu) ； 周茂振(Mao-Jhen Jhou)

《Journal of Data Analysis》 16卷2期 (2021/08) Pp. 43-57

https://doi.org/10.6338/JDA.202108_16(2).0003

摘要

隨著醫療科技的進步，每個人對於醫療品質的要求日益增加，如何精準預測癌症復發一直是醫療資訊領域重要的研究議題。機器學習技術在近年蓬勃發展，也已廣泛的應用於各種醫療資料分析的議題上。本研究建構多階段混合式分類架構於肝癌患者復發的預測研究中。在此預測架構中，將應用資料探勘中的羅吉斯迴歸、支援向量機、多元適性雲型迴歸和XGBoost為分類技術，並且考慮內嵌法與過濾法特徵選取技術、以及過採樣法與人工數據合成法處理資料不平衡技術以建構預測模式。本研究將所提之混合模式與單純模式的結果進行比較。實證結果顯示，無論資料切割比例，所提之混合預測模式的預測結果相較於單純預測模式較佳。並由最佳模式中可知，經由資料不平衡技術後再特徵選取能夠有效提升預測績效，並且所提模式能有效地建構肝癌復發的預測模式。

關鍵字

肝癌復發 ； 特徵選取 ； 資料不平衡 ； 混合預測模式

並列摘要

With the advancement of medical technology, everyone's requirements for medical quality are increasing. How to accurately predict cancer recurrence has always been an important research issue in the field of medical information. Machine learning technology has been widely used in various medical data analysis issues. This study constructs a multi-stage hybrid classification framework to predict the recurrence of liver cancer patients. In the proposed prediction model, the logistic regression, support vector machine, multivariate adaptive regression splines, and XGBoost methods are used as classification tools, the embedded method and filter method are used as feature selection tools, and the over-sampling method and synthetic data generation method are used to deal with data imbalance issues. This study compared the results of the proposed hybrid models with the single models. The empirical results showed that regardless of the data cutting ratio, the prediction results of the proposed hybrid predicting models are better than those of the single prediction models. This research provided a more suitable method for predicting liver cancer recurrence.

並列關鍵字

Liver cancer recurrence ； feature selection ； data imbalance ； hybrid prediction model

參考文獻

Chen, T., & Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. In the Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, 13-17 August, 785-794.

Fu, G. S., Levin-Schwartz, Y., Lin, Q. H., & Zhang, D. (2019). Machine Learning for Medical Imaging. J. Healthc. Eng.

Gu, X., Ni, T., & Wang, H. (2014). New fuzzy support vector machine for the class imbalance problem in medical datasets classification. The scientific world journal.

Kourou, K., Exarchos, T. P., Exarchos, K. P., Karamouzis, M. V., & Fotiadis, D. I. (2015). Machine learning applications in cancer prognosis and prediction. Comput. Struct. Biotechnol., 13, 8-17.

Leon, J., Namuche, F., Montenegro, P. C., & Flores, C. J. (2019). Risk factors predicting colorectal cancer recurrence in a Latin American population. J. Clin. Oncol., 37(15).