Claim data analysis

 Claim validation is one of the critical functional areas in the #pharma field.

The most basic approach would be based on supervised machine learning, in which you model claims concerning several existing categories based on existing approval processes. However, the quality of the training set is a separate issue that might be an even bigger problem.

Another basic approach would be based on feature extraction that can give us access to more complex constructs and treat those in a more explicit way as determining claims factors.

This paper also goes into interesting temporal issues, #bias, and sample statistics. Indeed, claims might be treated as time-series data that provides an additional dimension.

#sampling #demographic #efficacy #timeseries #regression #datascience #anova #covariance #bigdata #machinelearning #statistics #artificialintelligence #ai #data #research #healthcare

https://www.tandfonline.com/doi/full/10.3111/13696998.2015.1108919