Policy Survival Application
For predictive analysis, an introduction of a new “Policy Survival” application to study cohorts of policy contracts and its survival rate over time from the policy issued date to termination date. Survival curves are calculated for a defined range of issue and study dates and any defined policy termination reason at the month, quarter, and year levels. Confidence intervals are calculated using Greenwood’s formula.
The Insight Enterprise can now accommodate the new disability income IDEC 2013 rates which include a much broader range of risk variables as well as a 5 year monthly rate period. The risk variables have been extended from the CIDA structure of gender, occupation class, waiting period and claim cause to additionally include contract and market type, smoker, benefit period group, COLA status and diagnosis severity.
This menu-driven interface has been additionally enhanced. A “Source Status” tab was added to allow dynamical audits of the Source database tables providing date ranges and record count details. The “job execution” tab now monitors job progress at the more detailed SSIS package level. For multi-instances, the “active” instance is identified through the introduction of an iwMaster database. Multi-level security groups for users have also been added.
Job Steps/Package re-structure
The revised job steps/package structure allows for greater flexibility to enable or disable the cube processing and other processes. This is achieved by redefining the LoadType parameter values of the primary and secondary loads and adding the system table dbo.SSISPackagesForLoadType to define the job level in addition to the dbo.SSISPackages table.
System Status Audit Job
An audit job has been introduced to capture useful information of the server and database environment status. The audit job can be executed after a data load process to capture load and table statistical details useful to debug load issues and to analyze the volume of data loaded into the DW and cubes.
The system audits functions have been revised and extended. Fact record count measures added to allow reconciliation of cube results with the DW in the data flow audits. Integrity audits now identify records to be dropped or loaded into the DW, and if loaded, how it will be handled in the system. Table Type attribute has been added to allow better filtering of audit reports. For ease of presentation, the source audit cube has been split into two separate cubes, by volume and by integrity audits. In the data blow audits, version audits have been extended to identify new and terminated records, versions and non-status change versions.