Abstract：Objective To analyze the costs of hospitalization of patients with lung cancer to understand the influencing factors of lung cancer hospitalization costs, and to provide a relatively consummate evaluation method for lung cancer cost studies. Methods By Mathlab 7.1, we took age group, occupation and other factors as the input neurons, hospitalization costs the output neuron to simulate the analog neural network model, while sensitivity of the derived variables to the neural network model was used to evaluate the factors of hospitalization costs effect. At the same time, we used SAS 8.2 to build the logistic regression model, and made a comparison between two model results. Results The two model results showed that the top three factors of the hospital cost impact were number of days of hospitalization, drugs, and discharge outcome of the situation. Conclusion BP neural network model and logistic regression can be well applied to the analysis of influencing factors of hospitalization costs, but because of the BP neural network model has a wider range of conditions of application, more in-depth information mining and full depth analysis of the influencing factors, and thus has a better development and utilization prospects.
孙霖，祁爱琴，徐天和，高永. BP神经网络模型与logistic回归方法在肺癌病人住院费用影响因素分析中的比较[J]. 中国医院统计, 2015, 22(3): 173-175.
Sun Lin, Qi Aiqin, Xu Tianhe, Gao Yong. Comparison of BP neural network with logistic regression in influencing factor analysis of lung cancer hospitalization costs. journal1, 2015, 22(3): 173-175.
