查詢結果分析
來源資料
相關文獻
- 建構整合類神經網路與分類迴歸樹在大腸癌病患淋巴轉移之預測
- 整合類神經網路與無母數統計鑑別模式之二階段信用評等模式
- 整合類神經網路與分類迴歸樹在建構企業危機診斷模式上之應用
- 以類神經網路及分類迴歸樹輔助肝癌病患預測存活情形
- 資料探勘中集群模式與分類模式之建構--模糊自適應共振理論網路、分類迴歸樹與類神經網路之整合與應用
- 應用類神經網路與分類迴歸樹於肝癌分類模式
- Neural Network Procedures for Taguchi's Dynamic Problems
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
頁籤選單縮合
題名 | 建構整合類神經網路與分類迴歸樹在大腸癌病患淋巴轉移之預測=Construction of the Integrated Neural Network and "Classification and Regression Tree Approach" to Predict the Lymphatic Metastasis in Colorectal Cancer Patients |
---|---|
作者姓名(中文) | 張偉斌; 劉德明; 王照元; 紀櫻珍; 吳忠敏; 王堉璿; 張偉嶠; | 書刊名 | 北市醫學雜誌 |
卷期 | 8:1 2011.03[民100.03] |
頁次 | 頁13-22 |
分類號 | 416.245 |
關鍵詞 | 淋巴轉移; 類神經網路; 分類迴歸樹; Lymphatic metastasis; Artificial neural network; Classification and regression tree; |
語文 | 中文(Chinese) |
中文摘要 | 本研究主要結合兩種常用的人工智慧方式,分析8個變數來建構決策樹,以建立大腸癌患者淋巴是否會有轉移現象之預測模式,提供醫療人員臨床診斷、照護之參考。方法:本文以類神經網路整合分類迴歸樹演算法,建構出有關大腸癌患者淋巴是否轉移的預測模式。研究樣本為2010年高雄醫學大學40名已罹患大腸癌患者的相關檢驗數值,其中已轉移至淋巴筆數為25筆。先以類神經網路找出重要變數,再使用分類迴歸樹進行建構預測模式。結果:類神經網路挑選8個變數用來建立決策樹,以此種方式所建立的分類迴歸樹敏感度75%優於單獨使用分類迴歸樹(敏感度64.3%)或單獨使用類神經網路(敏感度60.8%)建構,經由這些變項可提供大腸癌淋巴轉移的預測模式。結論:本研究希望利用具有區分力的屬性建構模式,以提高模式的分類效能。 |
英文摘要 | The objective of this paper is to construct a prediction model for colorectal cancer patients with lymphatic metastasis. We hoped that this will help to provide a care reference model that has clinical application. Methods: We integrated a neural network and decision tree approach to build our predictive model. We collected information on patients from Kaoshiung Medical University Hospital who had suffered from colorectal cancer and investigated all relevant laboratory and genetic data. Out of 40 samples, 25 showed lymphatic metastasis. As a starting point, we adopted a neural network approach to pinpoint the important variables and then used a decision tree to construct a predictive model. Results: The neural network was used as a filter and identified eight variables that were then used to build the predictive model. We combined CART methodology and an artificial neural network to do this. Using the validation data set, the combined model had an accuracy of 83.3% and a sensitivity of 75%. Using the CART Model, the sensitivity was 63%, while using the ANN Model had a sensitivity of 60.8%. Conclusion: The variables identified using the approach described here was able to build a model that is able to improve the prediction of lymphatic metastasis. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。