Heterogeneous Mixture Learning Technology
Discovery of specific regularities in a variety of different data
Heterogeneous Mixture Learning Technology learns multiple relationships hidden in big data, discovers useful patterns and regularities, and selects the appropriate one depending on the situation. This enables higher-precision prediction in dynamically changing environments than existing machine learning technologies, which often take just a single pattern into account.
- Demand forecasting: Order placement operation at a retail store
- Fraudulent account detection: Detecting fraudulent transaction accounts
- Loan screening: Loan examination work