医疗与健康的数据分析与决策

Based on the special acquirement of decision demand in the field of healthcare, we try to find out the behavioral preference and pattern of patients, discover the characteristics and trends of healthcare providers, and finish the segment of the market. Then, we establish an evaluation system of healthcare service quality about patient satisfaction. By using DEA method, we evaluate the scale efficiency. Additionally, we investigate the fairness issue of healthcare resource allocation when patients make payment through different ways. Based on such studies, we combine the data of healthcare quality, efficiency and fairness. Then, we discuss healthcare data resource integration and scheduling technique based on cloud computing, and establish patient behavior-oriented multi-dimensional healthcare data warehouse under the cloud computing platform. Furthermore, using sparse least squares support vector machine, we study the comprehensive evaluation of healthcare risk. Moreover, taking the current sensitive doctor-patient conflicts as a starting point, we establish healthcare ethics risk monitoring model based on decision tree and neural networks and excavate healthcare fraud in the healthcare system. We use Markov and BP neural networks to forecast healthcare resource demand, in order to optimize online healthcare resource allocation. Our research will provide intellectual support for healthcare services agency when make intelligence decisions.

对医疗健康服务领域的决策进行需求调研,发现患者行为偏好、行为模式，服务提供商服务的特性、趋势，并对市场进行细分。建立基于患者满意度的医疗服务质量评价体系，利用动态数据包络分析（Data Envelopment Analysis，DEA）方法分析医疗服务规模效率，藉此完成度量医疗服务质量、效率与公平的数据集成。在上述研究基础上，讨论基于云计算的医疗健康数据资源整合与调度技术，并建立云计算平台下患者行为导向的多维医疗与健康数据仓库。进一步地，基于稀疏最小二乘支持向量机理论讨论医疗风险的综合评价问题，并以当前较敏感的医患矛盾为切入点，建立基于决策树和神经网络的医德风险监测模型，并挖掘分析医疗保障体系中的欺诈现象。利用Markov模型和BP神经网络对医疗资源需求进行预测，研究在线优化医疗和健康服务商的资源配置问题，为医疗健康机构的智能化决策提供智力支持。