Back in the 1990s, the financial services sector adopted generalized linear models (GLMs) because of their accuracy and the increasing complexity of decision-making. But GLMs suffer from several disadvantages. Specifically, they can be resource-intensive and time-consuming to build. They cannot work with missing values, and they require data to be collected from rigorously conducted experiments.
Enter automated machine learning and DataRobot’s automated machine learning platform. Our solution not only surmounts these obstacles but also provides more accurate, human-friendly explanations for how it uses the data, the patterns found in the data, and the reasons for a specific decision or prediction.
Read our white paper Advancing Financial Services Models Beyond GLMs , where we cover such topics as:
- What is a GLM and how this modeling technique was not designed for the banking or insurance industries
- How automated machine learning deals with missing values while also handling outliers in data
- How machine learning deals with the non-linear impact of socio-economic features that are an important driver of both financial risk and marketing activity