A multifeature statistical image segmentation algorithm is described. Multiple features such as grey, edge magnitude and correlation are combined to form a multidimensional space statistics. The statistical algorithm ...A multifeature statistical image segmentation algorithm is described. Multiple features such as grey, edge magnitude and correlation are combined to form a multidimensional space statistics. The statistical algorithm is used to segment an image using the decision curved surface determined by the multidimensional feature function. The segmentation problem which is difficult to solve using the features independently will be readily solved using the same features jointly. An adaptive segmentation algorithm is discussed. Test results of the real-time TV tracker newly developed have shown that the segmentation algorithm discussed here improves effectively the image segmentation quality and system tracking performance.展开更多
目的:构建基于机器学习的结核性胸膜炎诊断预测模型,提高临床诊断准确性。方法:回顾性收集2020年1月至2021年12月期间西安市胸科医院收治的523例胸腔积液患者(结核性胸膜炎375例,非结核性胸膜炎148例)的临床资料。纳入腺苷脱氨酶(adenos...目的:构建基于机器学习的结核性胸膜炎诊断预测模型,提高临床诊断准确性。方法:回顾性收集2020年1月至2021年12月期间西安市胸科医院收治的523例胸腔积液患者(结核性胸膜炎375例,非结核性胸膜炎148例)的临床资料。纳入腺苷脱氨酶(adenosine deaminase,ADA)、结核感染T细胞斑点试验(T-SPOT.TB)、C反应蛋白(C-reactive protein,CRP)等15项指标,采用随机森林、支持向量机、神经网络等7种机器学习算法构建预测模型,通过5折交叉验证评估模型性能,使用SHapley加法解释(SHapley Additive exPlanations,SHAP)算法进行特征重要性分析。结果:神经网络模型性能最优,测试集曲线下面积(area under the curve,AUC)为0.932,准确率为88.6%,精确率和召回率分别为94.4%和89.3%。SHAP分析显示,ADA(SHAP值为0.12~0.18)和T-SPOT.TB(SHAP值为0.10~0.15)是最重要的预测因子,且两者存在显著协同效应(P<0.001)。结论:本研究构建的神经网络模型具有较高的诊断效能,通过可解释性分析明确了关键预测因子及其交互作用,为结核性胸膜炎的精准诊断提供了新工具。该模型可辅助临床决策,特别适用于传统诊断中的“灰色区域”病例。展开更多
A multiple model tracking algorithm based on neural network and multiple-process noise soft-switching for maneuvering targets is presented.In this algorithm, the＂current＂statistical model and neural network are runn...A multiple model tracking algorithm based on neural network and multiple-process noise soft-switching for maneuvering targets is presented.In this algorithm, the＂current＂statistical model and neural network are running in parallel.The neural network algorithm is used to modify the adaptive noise filtering algorithm based on the mean value and variance of the＂current＂statistical model for maneuvering targets, and then the multiple model tracking algorithm of the multiple processing switch is used to improve the precision of tracking maneuvering targets.The modified algorithm is proved to be effective by simulation.展开更多
Based on the sequential probability ratio test(SPRT)developed by Wald,an improved method for successful probability test of missile flight is proposed.A recursive algorithm and its program in Matlab are designed to ca...Based on the sequential probability ratio test(SPRT)developed by Wald,an improved method for successful probability test of missile flight is proposed.A recursive algorithm and its program in Matlab are designed to calculate the real risk level of the sequential test decision and the average number of samples under various test conditions.A concept,that is "rejecting as soon as possible",is put forward and an alternate operation strategy is conducted.The simulation results show that it can reduce the test expenses.展开更多
文摘A multifeature statistical image segmentation algorithm is described. Multiple features such as grey, edge magnitude and correlation are combined to form a multidimensional space statistics. The statistical algorithm is used to segment an image using the decision curved surface determined by the multidimensional feature function. The segmentation problem which is difficult to solve using the features independently will be readily solved using the same features jointly. An adaptive segmentation algorithm is discussed. Test results of the real-time TV tracker newly developed have shown that the segmentation algorithm discussed here improves effectively the image segmentation quality and system tracking performance.
文摘目的:构建基于机器学习的结核性胸膜炎诊断预测模型,提高临床诊断准确性。方法:回顾性收集2020年1月至2021年12月期间西安市胸科医院收治的523例胸腔积液患者(结核性胸膜炎375例,非结核性胸膜炎148例)的临床资料。纳入腺苷脱氨酶(adenosine deaminase,ADA)、结核感染T细胞斑点试验(T-SPOT.TB)、C反应蛋白(C-reactive protein,CRP)等15项指标,采用随机森林、支持向量机、神经网络等7种机器学习算法构建预测模型,通过5折交叉验证评估模型性能,使用SHapley加法解释(SHapley Additive exPlanations,SHAP)算法进行特征重要性分析。结果:神经网络模型性能最优,测试集曲线下面积(area under the curve,AUC)为0.932,准确率为88.6%,精确率和召回率分别为94.4%和89.3%。SHAP分析显示,ADA(SHAP值为0.12~0.18)和T-SPOT.TB(SHAP值为0.10~0.15)是最重要的预测因子,且两者存在显著协同效应(P<0.001)。结论:本研究构建的神经网络模型具有较高的诊断效能,通过可解释性分析明确了关键预测因子及其交互作用,为结核性胸膜炎的精准诊断提供了新工具。该模型可辅助临床决策,特别适用于传统诊断中的“灰色区域”病例。
文摘A multiple model tracking algorithm based on neural network and multiple-process noise soft-switching for maneuvering targets is presented.In this algorithm, the＂current＂statistical model and neural network are running in parallel.The neural network algorithm is used to modify the adaptive noise filtering algorithm based on the mean value and variance of the＂current＂statistical model for maneuvering targets, and then the multiple model tracking algorithm of the multiple processing switch is used to improve the precision of tracking maneuvering targets.The modified algorithm is proved to be effective by simulation.
文摘Based on the sequential probability ratio test(SPRT)developed by Wald,an improved method for successful probability test of missile flight is proposed.A recursive algorithm and its program in Matlab are designed to calculate the real risk level of the sequential test decision and the average number of samples under various test conditions.A concept,that is "rejecting as soon as possible",is put forward and an alternate operation strategy is conducted.The simulation results show that it can reduce the test expenses.
