数据挖掘预测模型在脑损伤患者认知功能康复中的应用

摘要

To better predict the influencing factors of cognitive rehabilitation of acquired brain injury ( ABI) patients, we propose three prediction models which are based on decision tree (DT), multilayer perceptron (MLP) and general regression neural network (GRNN) respectively.By means of 10-fold cross validation test algorithm, we test the performance of the model by analysing its specificity, sensitivity and precision as well as the confusion matrix so as to obtain new knowledge to evaluate and improve the effectiveness of the cognitive rehabilitation process.Experimental results show that the simulation results based on DT model are clearly superior to other models, the averageprediction accuracy reaches up to 90.38%.%为了更好地预测后天性脑损伤ABI（ Acquired Brain Injury）患者认知功能康复的影响因素，提出基于决策树（ DT）、多层感知器（ MLP）和广义回归神经网络（ GRNN）的三种预测模型。借助于10折交叉验证测试算法，通过专一性、灵敏度和精度分析以及混淆矩阵分析对模型的性能进行测试，从而获得新的知识以评估和改善认知功能康复过程中的有效性。实验结果表明，基于DT的模型的模拟结果明显比其他模型更为优越，预测平均精度可高达90．38％。