gravatar

ogonzo1993

Omar Gonzalez

Recently Published

AppRISE: Loss Ratios by COB V2.1
This RMarkdown document will contain analysis and tables of Total Loss Ratios (VRUInc/CATPrem+AOPPrem) as well as CAT Loss Ratios (VRUInc/CATPrem) by Class of Business (COB). Note: most tables are interactive, so please feel free to toggle with columns to sort in ascending/descending order, apply filters and/or ranges, export, etc.
AppRISE: Model Miss Analysis (Model Predict - Actual)
Model Miss: AIREst - VRUInc Model Miss & RMSEst - VRUInc By storm. Caption will contain: (a) # instances of no loss and # instances with loss (b) aggregate `sum(model prediction - VRUInc)` [e.g., obs 1: RMSEst = 100 and VRUInc = 90... model miss = +10; obs 2: RMSEst = 200 and VRUInc = 500... model miss = -300. Aggregate miss = +10 + (-300) = -290] (c) Root Mean Square Error (d) Largest model miss (largest absolute value) **Note:** If figures are negative (**-**), AIREst or RMSEst underestimated (i.e., VRU incurred loss was higher than predicted). If figures are positive (**+**), AIREst or RMSEst is overestimated (i.e., VRU incurred loss was lower than predicted).
AppRISExVRU EDA