臺灣博碩士論文加值系統

目的 : 使用監督式機器學習演算法篩選頭頸癌患者接受放射治療後甲狀腺功能低下之預測因子。
材料與方法 : 本研究收集160位頭頸癌患者之回溯性資料，排除20位在接受放射線治療前已有甲狀腺功能低下或其它甲狀腺疾病之患者以及3位具劑量離群值的患者後，針對137位頭頸癌患者之資料使用監督式機器學習演算法進行預測因子之評估。候選因子包含性別、年齡、甲狀腺體積、最小劑量、平均劑量、最大劑量、治療次數、器官接受X劑量下所佔的相對體積 (X: 10, 20, 30, 40, 50, 60 Gy)。監督式機器學習演算法則採用決策樹 (Decision tree) 、隨機森林 (Random forest) 、支持向量機 (Support vector machine, SVM) 三種分類演算法，個別進行10次交叉驗證分析甲狀腺受損狀態及挑選甲狀腺功能受損之預測因子。另外並透過二元邏輯回歸分析 (Binary logistic regression)、皮爾森積差相關係數 (Pearson product-moment correlation coefficient) 以及最小絕對壓縮機制 (Least absolute shrinkage and selection operator, LASSO) 與三種機器學習演算法分析得到之預測因子進行重要性匹配，作為提高解釋性以及最佳化演算法之依據。透過分析上述三種預測甲狀腺併發症機率之演算法，得出相對應的準確性(Accuracy, ACC)、受試者特性曲線 (Receiver operating curve, ROC)、受試者特性曲線下面積(Area under the curve, AUC) 作為調整演算法參數與因子權重之參考依據。
結果 : 經由對三種監督式機器學習演算法預測因子進行重要性排序，其前五項預測因子分別為年齡、甲狀腺體積、平均劑量、V50與V60。其中年齡與體積因子為負相關，表示年齡與甲狀腺體積越大其甲狀腺損害風險越低；平均劑量、V50與V60為正相關，表示平均劑量、V50與V60越大其甲狀腺損害風險越高。對三種演算法預測甲狀腺損害之機率進行參數之優化結果如下: 決策樹演算法從AUC: 0.653、ACC: 70.6% 變更為AUC: 0.635、ACC: 73.5%；隨機森林演算法從AUC: 0.793、ACC: 79.4% 變更為AUC: 0.827、ACC: 82.4%；SVM演算法從AUC: 0.649、ACC: 79.4% 變更為AUC: 0.692、ACC: 79.4% 。其中以隨機森林演算法預測甲狀腺併發症之機率較為準確。
結論 :
研究發現於接受放射治療的頭頸癌患者中有五項預測因子(年齡、甲狀腺體積、平均劑量、V50與V60)，是影響發生甲狀腺功能低下機率之重要因素。透過監督式機器學習演算法之參數優化可以進一步提高預測頭頸癌患者接受放射治療後發生甲狀腺併發症機率之準確性。本研究之演算法可作為未來預測放射治療併發症機率與輔助醫療決策之參考依據。

Purpose: To assess the predictive factors for developing hypothyroidism in patients with head and neck cancer treated with radiotherapy by using supervised machine learning algorithms.
Materials and methods: This retrospective study collected clinical and dosimetric data of 160 head and neck cancer patients treated with radiotherapy in a medical center. Twenty patients with past history of hypothyroidism or other thyroid diseases before receiving radiotherapy, and 3 patients with dosimetric outlier were excluded. The clinical characteristics and dosimetric parameters of the remaining 137 patients were analyzed to assess the predictive factors for developing hypothyroidism by using supervised machine learning algorithms. Candidate factors include gender, age, thyroid volume, minimum dose, average dose, maximum dose, fraction number, relative volume of organs under X dose (X: 10, 20, 30, 40, 50, 60 Gy). The supervised machine learning algorithms include the followings: Decision tree, Random forest, and Support vector machine (SVM). Cross-validation analysis was conducted for ten times to identify the predictive factors for developing hypothyroidism. In addition, in order to improve interpretability and optimize the supervised machine learning algorithms, Binary logistic regression, Pearson product-moment correlation coefficient, and Least absolute shrinkage and selection operator (LASSO) were employed for further evaluation of the predictive factors identified by the supervised machine learning algorithms. Accuracy (ACC) and Receiver operating curve (ROC), Area under the curve (AUC) were used as a reference basis for adjusting algorithm parameters and factor weighting.
Results: Through ranking the importance of the predictive factors identified by the three supervised machine learning algorithms, the first five predictive factors are age, thyroid volume, mean dose, V50 and V60. Age and thyroid volume are negatively correlated with the risk of developing hypothyroidism. It means that the greater the age and the thyroid volume, the lower the risk of thyroid damage. The mean dose, V50 and V60 are positively correlated, indicating that the larger the mean dose, V50 and V60, the higher the risk of thyroid damage. The results of optimization of the parameters for predicting the probability of thyroid damage by the three algorithms are as follows: decision tree algorithm: AUC= 0.653, ACC= 70.6% changed to AUC= 0.635, ACC=73.5%; random forest algorithm: AUC= 0.793, ACC= 79.4% changed to AUC= 0.827, ACC= 82.4%; SVM algorithm: AUC=0.649, ACC=79.4% changed to AUC=0.692, ACC= 79.4%. Among these, random forest algorithm is associated with relatively higher accuracy in predicting thyroid complications.
Conclusions: The study revealed five predictors (age, thyroid volume, mean dose, V50 and V60) for developing radiation-related hypothyroidism. Through optimization of the parameters of the supervised machine learning algorithms, the accuracy of predicting the probability of radiation-related hypothyroidism for head and neck cancer patients could be further improved. The algorithms of this study can be used as a reference for predicting radiation-related complications and medical decision making in the future.

摘要
Abstract
致謝
目錄
表目錄
圖目錄
符號縮寫
符號說明
第一章 緒論
1.1 動機
1.2 目的
1.3 相關文獻探討
第二章 材料與方法
2.1 前言
2.2 患者資料
2.3 治療儀器以及治療計畫系統
2.4 甲狀腺損害風險之演算法
2.4.1 決策樹演算法
2.4.1.1 辨識甲狀腺損害風險
2.4.1.2 決策樹原理
2.4.1.3 決策樹演算流程
2.4.2 隨機森林演算法
2.4.3 支持向量機演算法
2.5 甲狀腺損害風險之分析與統計指標評估
2.5.1 分析過程
2.5.2 統計指標評估與描述
2.5.2.1 二元邏輯回歸分析
2.5.2.2 皮爾森積差相關係數
2.5.2.3 最小絕對壓縮機制
2.5.3 評估演算法
2.5.3.1 評估演算法之相關指標
2.5.3.2 繪製特性曲線ROC與曲線下面積AUC評估分數
2.6 最佳化甲狀腺損害風險演算法之參數調整
2.7 最佳化預測甲狀腺損害風險之能力
第三章 結果
3.1 前言
3.2 甲狀腺損害風險之分析與統計指標評估
3.2.1 二元邏輯回歸分析
3.2.2 皮爾森積差相關係數
3.2.3 最小絕對壓縮機制
3.3 甲狀腺損害風險演算法之初步訓練與測試結果
3.3.1 挑選預測因子重要性排序與統計指標匹配比較
3.3.2 未調整參數演算法之評估指標比較
3.3.3 未調整參數之特性曲線ROC與曲線下面積AUC評估比較
3.4 最佳化甲狀腺損害風險演算法之參數調整
3.4.1 決策樹演算法最佳化
3.4.2 隨機森林演算法最佳化
3.4.3 支持向量機演算法最佳化
3.5 最佳化預測甲狀腺損害風險之能力評估
3.5.1 最佳化演算法之特性曲線ROC與曲線下面積AUC評估比較
3.5.2 最佳化演算法與初步演算法優劣差距比較
第四章 討論
4.1 相關因子比較
4.2 其他演算法與機器學習比較
第五章 結論
參考文獻

1.Taiwan Cancer Registry Annual Report, Health Promotion Administration, Ministry of Health and Welfare. 2016.
2.Gomez-Millan, J., J.R. Fernández, and J.A.M. Carmona, Current status of IMRT in head and neck cancer. Reports of Practical Oncology & Radiotherapy, 2013. 18(6): p. 371-375.
3.Gupta, T., et al., Three-dimensional conformal radiotherapy (3D-CRT) versus intensity modulated radiation therapy (IMRT) in squamous cell carcinoma of the head and neck: a randomized controlled trial. Radiotherapy and Oncology, 2012. 104(3): p. 343-348.
4.Hanley, O. and M. Leech, Reduction of xerostomia in head and neck cancer patients. A critical review of the literature. Radiography, 2016. 22: p. S57-S63.
5.Liu, H., et al., Evaluation of 3D-CRT, IMRT and VMAT radiotherapy plans for left breast cancer based on clinical dosimetric study. Computerized Medical Imaging and Graphics, 2016. 54: p. 1-5.
6.Garden, A.S., et al. Target coverage for head and neck cancers treated with IMRT: review of clinical experiences. in Seminars in Radiation Oncology. 2004. Elsevier.
7.Gupta, T. and C.A. Narayan, Image-guided radiation therapy: Physician's perspectives. Journal of medical physics/Association of Medical Physicists of India, 2012. 37(4): p. 174.
8.Jereczek-Fossa, B.A., et al., Radiotherapy-induced thyroid disorders. Cancer treatment reviews, 2004. 30(4): p. 369-384.
9.Rønjom, M.F., Radiation-induced hypothyroidism after treatment of head and neck cancer. Dan Med J, 2016. 63(3): p. B5213.
10.Srikantia, N., et al., How common is hypothyroidism after external radiotherapy to neck in head and neck cancer patients? Indian journal of medical and paediatric oncology: official journal of Indian Society of Medical & Paediatric Oncology, 2011. 32(3): p. 143.
11.Grande, C., Hypothyroidism following radiotherapy for head and neck cancer: multivariate analysis of risk factors. Radiotherapy and Oncology, 1992. 25(1): p. 31-36.
12.Hancock, S.L., R.S. Cox, and I.R. McDougall, Thyroid diseases after treatment of Hodgkin's disease. New England Journal of Medicine, 1991. 325(9): p. 599-605.
13.Kamal, M., et al., Radiation-induced hypothyroidism after radical intensity modulated radiation therapy for oropharyngeal carcinoma. Advances in radiation oncology, 2020. 5(1): p. 111-119.
14.Kuten, A., et al., Postradiotherapy hypothyroidism: radiation dose response and chemotherapeutic radiosensitization at less than 40 Gy. Journal of surgical oncology, 1996. 61(4): p. 281-283.
15.Lee, V., et al., Dosimetric predictors of hypothyroidism after radical intensity-modulated radiation therapy for non-metastatic nasopharyngeal carcinoma. Clinical Oncology, 2016. 28(8): p. e52-e60.
16.Liening, D.A., et al., Hypothyroidism following radiotherapy for head and neck cancer. Otolaryngology—Head and Neck Surgery, 1990. 103(1): p. 10-13.
17.Lin, Z., et al., Evaluation of clinical hypothyroidism risk due to irradiation of thyroid and pituitary glands in radiotherapy of nasopharyngeal cancer patients. Journal of medical imaging and radiation oncology, 2013. 57(6): p. 713-718.
18.Mercado, G., et al., Hypothyroidism: a frequent event after radiotherapy and after radiotherapy with chemotherapy for patients with head and neck carcinoma. Cancer: Interdisciplinary International Journal of the American Cancer Society, 2001. 92(11): p. 2892-2897.
19.Rønjom, M.F., et al., Hypothyroidism after primary radiotherapy for head and neck squamous cell carcinoma: normal tissue complication probability modeling with latent time correction. Radiotherapy and Oncology, 2013. 109(2): p. 317-322.
20.Sinard, R.J., et al., Hypothyroidism after treatment for nonthyroid head and neck cancer. Archives of Otolaryngology–Head & Neck Surgery, 2000. 126(5): p. 652-657.
21.Weissler, M.C. and B.W. Berry, Thyroid‐stimulating hormone levels after radiotherapy and combined therapy for head and neck cancer. Head & neck, 1991. 13(5): p. 420-423.
22.Wu, Y.-H., et al., Hypothyroidism after radiotherapy for nasopharyngeal cancer patients. International Journal of Radiation Oncology* Biology* Physics, 2010. 76(4): p. 1133-1139.
23.Tell, R., et al., Long-term incidence of hypothyroidism after radiotherapy in patients with head-and-neck cancer. International Journal of Radiation Oncology* Biology* Physics, 2004. 60(2): p. 395-400.
24.Lyman, J.T., Complication probability as assessed from dose-volume histograms. Radiation Research, 1985. 104(2s): p. S13-S19.
25.Bakhshandeh, M., et al., Normal tissue complication probability modeling of radiation-induced hypothyroidism after head-and-neck radiation therapy. International Journal of Radiation Oncology* Biology* Physics, 2013. 85(2): p. 514-521.
26.Gulliford, S., Modelling of normal tissue complication probabilities (NTCP): review of application of machine learning in predicting NTCP, in Machine Learning in Radiation Oncology. 2015, Springer. p. 277-310.
27.Zhang, H., et al., SU‐HH‐AUD C‐03: Machine Learning Tools for Predicting Clinical Complications in a Multi‐Plan IMRT Framework. Medical Physics, 2008. 35(6Part18): p. 2854-2854.
28.Giraud, P., et al., Radiomics and machine learning for radiotherapy in head and neck cancers. Frontiers in oncology, 2019. 9: p. 174.
29.Abdollahi, H., et al., Cochlea CT radiomics predicts chemoradiotherapy induced sensorineural hearing loss in head and neck cancer patients: a machine learning and multi-variable modelling study. Physica Medica, 2018. 45: p. 192-197.
30.Bibault, J.-E., P. Giraud, and A. Burgun, Big data and machine learning in radiation oncology: state of the art and future prospects. Cancer letters, 2016. 382(1): p. 110-117.
31.Kang, J., et al., Machine learning approaches for predicting radiation therapy outcomes: a clinician's perspective. International Journal of Radiation Oncology* Biology* Physics, 2015. 93(5): p. 1127-1135.
32.Kearney, V., et al., The application of artificial intelligence in the IMRT planning process for head and neck cancer. Oral Oncology, 2018. 87: p. 111-116.
33.Fujiwara, M., et al., The threshold of hypothyroidism after radiation therapy for head and neck cancer: a retrospective analysis of 116 cases. Journal of radiation research, 2015. 56(3): p. 577-582.
34.Diaz, R., et al., Hypothyroidism as a consequence of intensity-modulated radiotherapy with concurrent taxane-based chemotherapy for locally advanced head-and-neck cancer. International Journal of Radiation Oncology* Biology* Physics, 2010. 77(2): p. 468-476.
35.Kim, E.S. and S.G. Yeo, Volumetric modulated arc radiotherapy sparing the thyroid gland for early‑stage glottic cancer: A dosimetrical analysis. Oncology Letters, 2014. 7(6): p. 1987-1991.
36.Alterio, D., et al., Thyroid disorders in patients treated with radiotherapy for head-and-neck cancer: a retrospective analysis of seventy-three patients. International Journal of Radiation Oncology* Biology* Physics, 2007. 67(1): p. 144-150.
37.Murthy, V., et al., Hypothyroidism after 3‐dimensional conformal radiotherapy and intensity‐modulated radiotherapy for head and neck cancers: Prospective data from 2 randomized controlled trials. Head & neck, 2014. 36(11): p. 1573-1580.
38.Vogelius, I.R., et al., Risk factors for radiation‐induced hypothyroidism: A literature‐based meta‐analysis. Cancer, 2011. 117(23): p. 5250-5260.
39.Pedregosa, F., et al., Scikit-learn: Machine learning in Python. the Journal of machine Learning research, 2011. 12: p. 2825-2830.
40.Chambers, J., Software for data analysis: programming with R. 2008: Springer Science & Business Media.
41.Moler, C.B., Numerical computing with MATLAB. 2004: SIAM.
42.McKinney, W., Python for data analysis: Data wrangling with Pandas, NumPy, and IPython. 2012: " O'Reilly Media, Inc.".
43.Tosi, S., Matplotlib for Python developers. 2009: Packt Publishing Ltd.
44.Walt, S.v.d., S.C. Colbert, and G. Varoquaux, The NumPy array: a structure for efficient numerical computation. Computing in science & engineering, 2011. 13(2): p. 22-30.
45.Hosmer Jr, D.W., S. Lemeshow, and R.X. Sturdivant, Applied logistic regression. Vol. 398. 2013: John Wiley & Sons.
46.Häne, B.G., K. Jäger, and H.G. Drexler, The Pearson product‐moment correlation coefficient is better suited for identification of DNA fingerprint profiles than band matching algorithms. Electrophoresis, 1993. 14(1): p. 967-972.
47.Colombani, C., et al., Application of Bayesian least absolute shrinkage and selection operator (LASSO) and BayesCπ methods for genomic selection in French Holstein and Montbéliarde breeds. Journal of Dairy Science, 2013. 96(1): p. 575-591.
48.Kong, C., et al., LASSO-based NTCP model for radiation-induced temporal lobe injury developing after intensity-modulated radiotherapy of nasopharyngeal carcinoma. Scientific reports, 2016. 6(1): p. 1-8.
49.Lee, T.-F., et al., Using multivariate regression model with least absolute shrinkage and selection operator (LASSO) to predict the incidence of xerostomia after intensity-modulated radiotherapy for head and neck cancer. PloS one, 2014. 9(2): p. e89700.
50.Lee, T.-F., et al., LASSO NTCP predictors for the incidence of xerostomia in patients with head and neck squamous cell carcinoma and nasopharyngeal carcinoma. Scientific reports, 2014. 4: p. 6217.
51.Bishop, C.M., Pattern recognition and machine learning. 2006: springer.
52.Mitchell, T.M., The discipline of machine learning. Vol. 9. 2006: Carnegie Mellon University, School of Computer Science, Machine Learning ….
53.Witten, I.H. and E. Frank, Data mining: practical machine learning tools and techniques with Java implementations. Acm Sigmod Record, 2002. 31(1): p. 76-77.
54.Niknejad, A. and D. Petrovic, Introduction to computational intelligence techniques and areas of their applications in medicine. Med Appl Artif Intell, 2013. 51.
55.Doshi-Velez, F. and B. Kim, Towards a rigorous science of interpretable machine learning. arXiv preprint arXiv:1702.08608, 2017.
56.García, S., A. Fernández, and F. Herrera, Enhancing the effectiveness and interpretability of decision tree and rule induction classifiers with evolutionary training set selection over imbalanced problems. Applied Soft Computing, 2009. 9(4): p. 1304-1314.
57.Gilpin, L.H., et al. Explaining explanations: An overview of interpretability of machine learning. in 2018 IEEE 5th International Conference on data science and advanced analytics (DSAA). 2018. IEEE.
58.Rudin, C., Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 2019. 1(5): p. 206-215.
59.Wang, G., et al., Sentiment classification: The contribution of ensemble learning. Decision support systems, 2014. 57: p. 77-93.
60.Webb, G.I. and Z. Zheng, Multistrategy ensemble learning: Reducing error by combining ensemble learning techniques. IEEE Transactions on Knowledge and Data Engineering, 2004. 16(8): p. 980-991.
61.Zhang, C. and Y. Ma, Ensemble machine learning: methods and applications. 2012: Springer.
62.Safavian, S.R. and D. Landgrebe, A survey of decision tree classifier methodology. IEEE transactions on systems, man, and cybernetics, 1991. 21(3): p. 660-674.
63.Tu, P.-L. and J.-Y. Chung. A new decision-tree classification algorithm for machine learning. in TAI'92-Proceedings Fourth International Conference on Tools with Artificial Intelligence. 1992. IEEE Computer Society.
64.Marjanović, M., et al., Landslide susceptibility assessment using SVM machine learning algorithm. Engineering Geology, 2011. 123(3): p. 225-234.
65.Shon, T. and J. Moon, A hybrid machine learning approach to network anomaly detection. Information Sciences, 2007. 177(18): p. 3799-3821.
66.Soman, K., R. Loganathan, and V. Ajay, Machine learning with SVM and other kernel methods. 2009: PHI Learning Pvt. Ltd.
67.Suykens, J.A. and J. Vandewalle, Least squares support vector machine classifiers. Neural processing letters, 1999. 9(3): p. 293-300.
68.Abraham, A., et al., Machine learning for neuroimaging with scikit-learn. Frontiers in neuroinformatics, 2014. 8: p. 14.
69.Komer, B., J. Bergstra, and C. Eliasmith. Hyperopt-sklearn: automatic hyperparameter configuration for scikit-learn. in ICML workshop on AutoML. 2014. Citeseer.
70.Nelli, F., Machine Learning with scikit-learn, in Python Data Analytics. 2018, Springer. p. 313-347.
71.Chatterjee, A., M. Vallières, and J. Seuntjens, Overlooked pitfalls in multi-class machine learning classification in radiation oncology and how to avoid them. Physica Medica, 2020. 70: p. 96-100.
72.Quinlan, J.R., Induction of decision trees. Machine learning, 1986. 1(1): p. 81-106.
73.Rutkowski, L., et al., The CART decision tree for mining data streams. Information Sciences, 2014. 266: p. 1-15.
74.Awad, W. and S. ELseuofi, Machine learning methods for spam e-mail classification. International Journal of Computer Science & Information Technology (IJCSIT), 2011. 3(1): p. 173-184.
75.Dannenberg, R.B., B. Thom, and D. Watson, A machine learning approach to musical style recognition. 1997.
76.Mohammed, A.A., et al., Human face recognition based on multidimensional PCA and extreme learning machine. Pattern Recognition, 2011. 44(10-11): p. 2588-2597.
77.Fürnkranz, J. and G. Widmer, Incremental reduced error pruning, in Machine Learning Proceedings 1994. 1994, Elsevier. p. 70-77.
78.Díaz-Uriarte, R. and S.A. De Andres, Gene selection and classification of microarray data using random forest. BMC bioinformatics, 2006. 7(1): p. 3.
79.Breiman, L., Random forests. Machine learning, 2001. 45(1): p. 5-32.
80.Oshiro, T.M., P.S. Perez, and J.A. Baranauskas. How many trees in a random forest? in International workshop on machine learning and data mining in pattern recognition. 2012. Springer.
81.Bickel, P.J. and D.A. Freedman, Asymptotic normality and the bootstrap in stratified sampling. The annals of statistics, 1984: p. 470-482.
82.Wichmann, F.A. and N.J. Hill, The psychometric function: II. Bootstrap-based confidence intervals and sampling. Perception & psychophysics, 2001. 63(8): p. 1314-1329.
83.Breiman, L., Classification and Regression Trees, Waldsworth International (1984); L. Breiman, Bagging Predictors. Machine Learning, 1996. 26: p. 123.
84.Tu, M.C., D. Shin, and D. Shin. Effective diagnosis of heart disease through bagging approach. in 2009 2nd International Conference on Biomedical Engineering and Informatics. 2009. IEEE.
85.Frohlich, H., O. Chapelle, and B. Scholkopf. Feature selection for support vector machines by means of genetic algorithm. in Proceedings. 15th IEEE International Conference on Tools with Artificial Intelligence. 2003. IEEE.
86.Joachims, T. Transductive inference for text classification using support vector machines. in Icml. 1999.
87.Pontil, M. and A. Verri, Support vector machines for 3D object recognition. IEEE transactions on pattern analysis and machine intelligence, 1998. 20(6): p. 637-646.
88.Scholkopf, B., et al., Comparing support vector machines with Gaussian kernels to radial basis function classifiers. IEEE transactions on Signal Processing, 1997. 45(11): p. 2758-2765.
89.Chen, C., et al., The diagnostic value of radiomics-based machine learning in predicting the grade of meningiomas using conventional magnetic resonance imaging: a preliminary study. Frontiers in oncology, 2019. 9: p. 1338.
90.Kniep, H.C., et al., Radiomics of brain MRI: utility in prediction of metastatic tumor type. Radiology, 2019. 290(2): p. 479-487.
91.Liao, X., et al., Machine‐learning based radiogenomics analysis of MRI features and metagenes in glioblastoma multiforme patients with different survival time. Journal of Cellular and Molecular Medicine, 2019. 23(6): p. 4375-4385.
92.Ortiz-Ramón, R., et al., Classifying brain metastases by their primary site of origin using a radiomics approach based on texture analysis: a feasibility study. European radiology, 2018. 28(11): p. 4514-4523.
93.Bradley, A.P., The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern recognition, 1997. 30(7): p. 1145-1159.
94.Ferri, C., J. Hernández-Orallo, and M.A. Salido. Volume under the ROC surface for multi-class problems. in European conference on machine learning. 2003. Springer.
95.Hand, D.J. and R.J. Till, A simple generalisation of the area under the ROC curve for multiple class classification problems. Machine learning, 2001. 45(2): p. 171-186.
96.Pencina, M.J., et al., Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Statistics in medicine, 2008. 27(2): p. 157-172.
97.Genovese, C.R., K. Roeder, and L. Wasserman, False discovery control with p-value weighting. Biometrika, 2006. 93(3): p. 509-524.
98.Bengio, Y. and Y. Grandvalet, No unbiased estimator of the variance of k-fold cross-validation. Journal of machine learning research, 2004. 5(Sep): p. 1089-1105.
99.Blum, A., A. Kalai, and J. Langford. Beating the hold-out: Bounds for k-fold and progressive cross-validation. in Proceedings of the twelfth annual conference on Computational learning theory. 1999.
100.Rodriguez, J.D., A. Perez, and J.A. Lozano, Sensitivity analysis of k-fold cross validation in prediction error estimation. IEEE transactions on pattern analysis and machine intelligence, 2009. 32(3): p. 569-575.
101.Bradley, J.K., et al., Parallel coordinate descent for l1-regularized loss minimization. arXiv preprint arXiv:1105.5379, 2011.
102.Graham, M. and J. Kennedy. Using curves to enhance parallel coordinate visualisations. in Proceedings on Seventh International Conference on Information Visualization, 2003. IV 2003. 2003. IEEE.
103.Richtárik, P. and M. Takáč, Parallel coordinate descent methods for big data optimization. Mathematical Programming, 2016. 156(1-2): p. 433-484.
104.Gotz, D., et al., Understanding care plans of community acquired pneumonia based on Sankey diagram. 2016.
105.Riehmann, P., M. Hanfler, and B. Froehlich. Interactive sankey diagrams. in IEEE Symposium on Information Visualization, 2005. INFOVIS 2005. 2005. IEEE.
106.Schmidt, M., The Sankey diagram in energy and material flow management: part II: methodology and current applications. Journal of industrial ecology, 2008. 12(2): p. 173-185.
107.Dawson, R., How significant is a boxplot outlier? Journal of Statistics Education, 2011. 19(2).
108.Schwertman, N.C., M.A. Owens, and R. Adnan, A simple more general boxplot method for identifying outliers. Computational statistics & data analysis, 2004. 47(1): p. 165-174.
109.Zani, S., M. Riani, and A. Corbellini, Robust bivariate boxplots and multiple outlier detection. Computational Statistics & Data Analysis, 1998. 28(3): p. 257-270.
110.Ogutu, J.O., T. Schulz-Streeck, and H.-P. Piepho. Genomic selection using regularized linear regression models: ridge regression, lasso, elastic net and their extensions. in BMC proceedings. 2012. Springer.
111.Massimino, M., et al., Thyroid-stimulating hormone suppression for protection against hypothyroidism due to craniospinal irradiation for childhood medulloblastoma/primitive neuroectodermal tumor. International Journal of Radiation Oncology* Biology* Physics, 2007. 69(2): p. 404-410.
112.Massimino, M., et al., TSH suppression as a possible means of protection against hypothyroidism after irradiation for childhood Hodgkins lymphoma. Pediatric blood & cancer, 2011. 57(1): p. 166-168.
113.Bantle, J.P., C.K. Lee, and S.H. Levitt, Thyroxine administration during radiation therapy to the neck does not prevent subsequent thyroid dysfunction. International Journal of Radiation Oncology* Biology* Physics, 1985. 11(11): p. 1999-2002.
114.Cicchetti, D.V., Neural networks and diagnosis in the clinical laboratory: state of the art. Clinical chemistry, 1992. 38(1): p. 9-10.
115.Cochran, A.J., Prediction of outcome for patients with cutaneous melanoma. Pigment cell research, 1997. 10(3): p. 162-167.
116.Cruz, J.A. and D.S. Wishart, Applications of machine learning in cancer prediction and prognosis. Cancer informatics, 2006. 2: p. 117693510600200030.
117.Exarchos, K.P., Y. Goletsis, and D.I. Fotiadis, Multiparametric decision support system for the prediction of oral cancer reoccurrence. IEEE Transactions on Information Technology in Biomedicine, 2011. 16(6): p. 1127-1134.
118.Kononenko, I., Machine learning for medical diagnosis: history, state of the art and perspective. Artificial Intelligence in medicine, 2001. 23(1): p. 89-109.
119.Sun, Y., et al., Improved breast cancer prognosis through the combination of clinical and genetic markers. Bioinformatics, 2007. 23(1): p. 30-37.
120.ScienceDirect.
121.Fielding, L.P., C.M. Fenoglio‐Preiser, and L.S. Freedman, The future of prognostic factors in outcome prediction for patients with cancer. Cancer, 1992. 70(9): p. 2367-2377.
122.Bach, P.B., et al., Variations in lung cancer risk among smokers. Journal of the National Cancer Institute, 2003. 95(6): p. 470-478.
123.Domchek, S.M., et al., Application of breast cancer risk prediction models in clinical practice. Journal of Clinical Oncology, 2003. 21(4): p. 593-601.
124.Gasco, F., et al., Childhood obesity and hormonal abnormalities associated with cancer risk. European journal of cancer prevention, 2004. 13(3): p. 193-197.
125.Ren, X., et al., ellipsoidFN: a tool for identifying a heterogeneous set of cancer biomarkers based on gene expressions. Nucleic acids research, 2013. 41(4): p. e53-e53.
126.Ren, X., et al., iPcc: a novel feature extraction method for accurate disease class discovery and prediction. Nucleic acids research, 2013. 41(14): p. e143-e143.
127.Wang, Y., et al., Revealing metabolite biomarkers for acupuncture treatment by linear programming based feature selection. BMC systems biology, 2012. 6(S1): p. S15.
128.Kim, W., et al., Development of novel breast cancer recurrence prediction model using support vector machine. Journal of breast cancer, 2012. 15(2): p. 230-238.
129.Ahmad, L.G., et al., Using three machine learning techniques for predicting breast cancer recurrence. J Health Med Inform, 2013. 4(124): p. 3.
130.Ayer, T., et al., Breast cancer risk estimation with artificial neural networks revisited: discrimination and calibration. Cancer, 2010. 116(14): p. 3310-3321.
131.Chang, S.-W., et al., Oral cancer prognosis based on clinicopathologic and genomic markers using a hybrid of feature selection and machine learning methods. BMC bioinformatics, 2013. 14(1): p. 170.
132.Chen, Y.-C., W.-C. Ke, and H.-W. Chiu, Risk classification of cancer survival using ANN with gene expression data from multiple laboratories. Computers in biology and medicine, 2014. 48: p. 1-7.
133.Delen, D., G. Walker, and A. Kadam, Predicting breast cancer survivability: a comparison of three data mining methods. Artificial intelligence in medicine, 2005. 34(2): p. 113-127.
134.Gevaert, O., et al., Predicting the prognosis of breast cancer by integrating clinical and microarray data with Bayesian networks. Bioinformatics, 2006. 22(14): p. e184-e190.
135.Kim, J. and H. Shin, Breast cancer survivability prediction using labeled, unlabeled, and pseudo-labeled patient data. Journal of the American Medical Informatics Association, 2013. 20(4): p. 613-618.
136.Listgarten, J., et al., Predictive models for breast cancer susceptibility from multiple single nucleotide polymorphisms. Clinical cancer research, 2004. 10(8): p. 2725-2737.
137.Park, C., et al., Integrative gene network construction to analyze cancer recurrence using semi-supervised learning. PloS one, 2014. 9(1): p. e86309.
138.Park, K., et al., Robust predictive model for evaluating breast cancer survivability. Engineering Applications of Artificial Intelligence, 2013. 26(9): p. 2194-2205.
139.Rosado, P., et al., Survival model in oral squamous cell carcinoma based on clinicopathological parameters, molecular markers and support vector machines. Expert systems with applications, 2013. 40(12): p. 4770-4776.
140.Stojadinovic, A., et al., Development of a Bayesian Belief Network Model for personalized prognostic risk assessment in colon carcinomatosis. The American Surgeon, 2011. 77(2): p. 221-230.
141.Tseng, C.-J., et al., Application of machine learning to predict the recurrence-proneness for cervical cancer. Neural Computing and Applications, 2014. 24(6): p. 1311-1316.
142.Waddell, M., D. Page, and J. Shaughnessy Jr. Predicting cancer susceptibility from single-nucleotide polymorphism data: a case study in multiple myeloma. in Proceedings of the 5th international workshop on Bioinformatics. 2005.
143.Xu, X., et al. A gene signature for breast cancer prognosis using support vector machine. in 2012 5th International Conference on BioMedical Engineering and Informatics. 2012. IEEE.
144.Chetvertkov, M.A., et al., Use of regularized principal component analysis to model anatomical changes during head and neck radiation therapy for treatment adaptation and response assessment. Medical physics, 2016. 43(10): p. 5307-5319.
145.Gomathy, V. and U. Snekhalatha. Automated segmentation using PCA and area estimation of thyroid gland using ultrasound images. in 2015 International Conference on Innovations in Information, Embedded and Communication Systems (ICIIECS). 2015. IEEE.