Metrics
3 File views/ downloads
32 Record Views
Thesis
Trend features and additive feature selection methods for churn models under property and casualty insurance business paradigms
University of West Florida Libraries
Master of Science (MS), University of West Florida
2018
Abstract
ABSTRACT: The P&C Small Commercial Insurance industry presents a signiαcant policy retention challenge given the presence of independent intermediaries. Current retention models generally follow traditional relational marketing paradigms and therefore do not account for the complexities introduced by the presence of an independent intermediary. Given an anonymous policy dataset consisting of correlated, high-dimensional, and high-cardinality categorical/numeric data; current data mining methods are used to construct a retention model with practical applications. Additionally, predictive features that capture intermediary-related information are engineered and designated as candidate features. Candidate features are selected for αnal model inclusion using various data mining approaches to feature importance measurement and feature selection.
Details
- Title
- Trend features and additive feature selection methods for churn models under property and casualty insurance business paradigms
- Resource Type
- Thesis
- Publisher
- University of West Florida Libraries; Pensacola, Florida :
- Format
- pdf
- Number of pages
- xi, 124 leaves
- Identifiers
- 99380090731106600
- Academic Unit
- Mathematics and Statistics
- Language
- English
- Awarding Institution
- University of West Florida; Master of Science (MS)
- Theses and Dissertations
- Master of Science (MS), University of West Florida