34145 Pacific Coast Hwy #335, Dana Point, CA 92629

Technical Features

  • 50 MB disk space required
  • Memory dependent upon size of dataset to be analyzed
  • CSV file import
  • Models Boolean Classifier and Regression problem types
  • Automatic feature (variable) reduction process for problems with large variable sets
  • Generates models in platform independent format, usable by Siearrabolics Model Runtime SDK
  • Binary Classification Models generate predictions with confidence
  • Uses the Sierrabolics ZGP analysis engine for data modeling
  • Generate and select models which function in the real world with the accuracy and efficacy as predicted during the analysis phase
  • Automatic variable type detection
  • Ability to import new data for evaluation by previously generated models as well as export the evaluated data
  • Full reporting of model characteristics, explanatory variables employed, and results of model applied to input dataset
  • Advanced functions:
    •       Real time Pareto efficiency Plot of all candidate solutions in analysis session for binary classification problems (TPR vs TNR).
    •       Real time Analysis performance graph
    •       Real time Model vs Prediction graph (Regression Problems)
    •       Real time Error Magnitude Kernel Density Estimation plot (Regression Problems)
    •       Real time Confidence Score Kernel Density Estimation plot (Binary Classification Problems)
    •       Real time ROC Curve plot (Binary Classification Problems)
    •       Detailed and comprehensive Binary Classification model statistics:
      •       True Positive Rate / Sensitivity
      •       True Negative Rate / Specificity
      •       False Positive Rate / Fall Out
      •       False Negative Rate
      •       Precision / Positive Predictive Value
      •       Negative Predictive Value
      •       False Discovery Rate
      •       False Omission Rate
      •       Positive Likelihood Ratio
      •       Negative Likelihood Ratio
      •       Diagnostic Odds Ratio
      •       Area Under ROC Curve
      •       F1 Score
      •       Matthews Correlation Coefficient
      •       Average Confidence Score for Correct Predictions
      •       Average Confidence Score for Incorrect Predictions