首页>
外国专利>
Method and apparatus for predicting whether a specified event will occur after a specified trigger event has occurred
Method and apparatus for predicting whether a specified event will occur after a specified trigger event has occurred
展开▼
机译:用于预测在指定触发事件发生之后是否将发生指定事件的方法和装置
展开▼
页面导航
摘要
著录项
相似文献
摘要
In many situations it is required to predict if and/or when an event will occur after a trigger. For example, businesses such as banks would like to predict if and when their customers are likely to leave after a particular event such as closing a loan. The business is then able to take action to prevent loss of customers. Customer data including data about customer who have closed a loan and then left a bank for example, is used to create a Bayesian statistical model. A plurality of attributes are available for each customer and the model involves partitioning these attributes into a plurality of partitions. In one embodiment the Bayesian statistical model is a survival analysis type model and in another embodiment the model comprises fitting a Weibull distribution to the data in each of the partitions. The marginal likelihood of the data is calculated and then the method involves mixing over all possible partitions in a Bayesian framework. Alternatively an optimal set of partitions which best predicts the data is chosen.
展开▼