Techniques for automatically detecting inactive addresses of providers using unsupervised learning approaches are provided. The techniques include determining an activity trend of an address. Responsive to determining that the address is associated with increasing activity, determining an active metric of the address, which indicates a likelihood of the address being active. Responsive to determining that the address is associated with decreasing activity, determining an inactive metric of the address, which indicates a likelihood of the address being inactive. The techniques further include determining whether the address is active or inactive based on the active metric and/or inactive metric. In some embodiments, the active metric or inactive metric is a weighted sum of z scores determined based on Gaussian distributions generated with respect to various benchmarks or provider features. In some embodiments, the weights used to determine the active metric or inactive metric are determined using RLHF techniques.

This disclosure relates generally to method and system for identifying frauds in unemployment insurance claims using hybrid weighted decision model. The method combines predictive power of machine learning models with domain knowledge infused key fraud indicators and network analysis to identify false positive claims. Initially, a set of claim information from an request of a claimant is extracted to assess risk affecting eligibility of the claimant to receive benefits. Further, a classification probability for a set of key fraud indicators are predicted for each claim using a set of top features associated with a prescient artificial intelligence (AI) model. Finally, one or more frauds associated with the unemployment insurance claim of the claimant are identified based on the weighted probability of each claim, a network diagram generated using the weighted probability for each claim, a set of filtered claims, and a set of business rules.