18443912. Computer-Implemented Method for Optimizing a Detection Threshold of a Prediction Model simplified abstract (Robert Bosch GmbH)

Computer-Implemented Method for Optimizing a Detection Threshold of a Prediction Model

Organization Name

Robert Bosch GmbH

Inventor(s)

Daniel Zander of Oberriexingen (DE)

Anton Iakovlev of Boeblingen (DE)

Damir Shakirov of Niefern-Oeschelbronn (DE)

David Schoenleber of Karlsruhe (DE)

Erin Sebastian Schmidt of Heimsheim (DE)

Jonas Bergdolt of Bietigheim-Bissingen (DE)

Jonathan Levin of Reutlingen (DE)

Matthias Werner of Stuttgart (DE)

Stefan Patrick Lindt of Ettlingen (DE)

Timo Pfrommer of Stuttgart (DE)

Uwe Lehmann of Dresden (DE)

Computer-Implemented Method for Optimizing a Detection Threshold of a Prediction Model - A simplified explanation of the abstract

This abstract first appeared for US patent application 18443912 titled 'Computer-Implemented Method for Optimizing a Detection Threshold of a Prediction Model

The abstract describes a computer-implemented method for optimizing the detection threshold of a prediction model used to identify anomalies in components.

• The method involves providing the prediction model and pre-test results for test parameters of components.
• Final test results indicating anomalies are provided for each component.
• Anomaly results are calculated by evaluating pre-test results with the prediction model.
• Explainability values for test parameters are calculated for each component.
• The detection threshold is optimized based on anomaly results, explainability values, and final test results.

• • Key Features and Innovation:**

• Optimization of detection threshold for anomaly detection in components.
• Utilization of pre-test results and explainability values for parameter optimization.
• Integration of final test results to enhance accuracy of anomaly detection.

• • Potential Applications:**

• Quality control in manufacturing processes.
• Fault detection in machinery and equipment.
• Cybersecurity for identifying anomalies in network behavior.

• • Problems Solved:**

• Enhances accuracy in identifying anomalies in components.
• Improves efficiency in anomaly detection processes.
• Provides a systematic approach to optimizing detection thresholds.

• • Benefits:**

• Increased precision in anomaly detection.
• Reduction in false positives and false negatives.
• Enhanced reliability in predicting component anomalies.

• • Commercial Applications:**

Optimizing Detection Thresholds for Anomaly Detection in Components: Market Implications

• • Questions about the Technology:**

1. How does this method improve anomaly detection compared to traditional approaches?

  This method improves anomaly detection by optimizing the detection threshold based on pre-test results, explainability values, and final test results, leading to increased accuracy and efficiency.

2. What are the potential challenges in implementing this method in real-world applications?

  Potential challenges in implementing this method may include data integration, model calibration, and validation processes to ensure optimal performance in diverse environments.

Original Abstract Submitted

A computer-implemented method for optimizing a detection threshold of a prediction model used to determine an anomaly of a component is disclosed. The detection threshold indicates the criterion above which the prediction model classifies a component as anomalous. The method includes (i) providing the prediction model, (ii) providing a plurality of pre-test results determined for a plurality of test parameters for a plurality of components, respectively, (iii) providing a final test result for each of the plurality of components, wherein the respective final test result indicates whether the respective component has an anomaly in a final test, (iv) calculating a respective anomaly result for each of the plurality of components by evaluating the respective pre-test results by the prediction model, (v) calculating a respective explainability value for each of the plurality of test parameters for each of the plurality of components by the prediction model, and (vi) optimizing the detection threshold based on the calculated anomaly results, the calculated explainability values and the final test results provided.