Dimont’s clients typically purchase portfolios of distressed mortgages and contract with Dimont to process the claims during the sales process. Dimont’s rich historical data provides the raw material needed to predict damage, claims, and recovery. Complexity in the data required a machine learning algorithm to understand better the collection of property attributes.
Anonymized Data from Dimont’s operational systems is deposited in S3. Amazon Key Management Service (KMS) along with security policies is used to guard against unwanted data access. Elastic Map Reduce (EMR) is used to scrub the data and identify features in the data. Features related to the property, geography and peril are extracted and transformed to create training data for the machine learning process.
Amazon Machine Learning service is used to create nine models. These models predict the probability of a claim, recovery and peril type. The models are adjusted to maximize prediction accuracy across a portfolio while evenly splitting false positive and false negative prediction errors.
To predict the claims associated with a portfolio of mortgages, the data file is dropped into an S3 bucket. EMR is used to scrub the data, identify features and append geography and peril attributes. The prediction models are applied to predict claims and recovery for the properties contained in the portfolio. The results are aggregated using EMR to provide a final report for the client investor.
These models are over 90% accurate in predicting existing damage and claims on a portfolio of loans. This service helps Dimont differentiate their service. Predicted claim performance allows investors to incorporate damage risk and recovery into their disposition decisions.