Free Informational Use Case Download: Using A Narrow Data Set to Accurately Predict Diabetes Model Predicts Diabetes Risk with 77% Accuracy — Without Reliance on Historical Indicators

This real-world scientific study leveraged the proprietary data lake alongside machine learning capabilities to answer pressing questions regarding diabetes diagnosis and population health management, such as:

  • Would it be possible to produce highly accurate predictive results using a narrow set of healthcare information?
  • Would it be possible to predict diabetes risk based solely on unrelated historical test results?

This scientific use case uncovers the background, design, and data that went into this study and highlights possible areas for future innovation. The use case also digs deep into the results of the study, including a 77% diabetes prediction rate, without reliance on traditional indicators such as glucose.