Supervised Machine Learning Validation for Target Failure Probability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing supervised machine learning methods lack reliability for safety-critical applications, as they fail to provide sufficient evidence of safety standards during operation, particularly in validating the performance of algorithms trained using artificial intelligence.
Innovation Solution
A method for supervised machine learning that involves creating a data pool with input and output data sets, dividing them into training and validation sets, training the algorithm using varied parameters, and validating the results to ensure a safety factor-related target probability of failure is not exceeded, thereby ensuring high reliability and safety standards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine learning algorithms are used to optimize safety-relevant applications, then the productivity and adaptability of the system are improved, but the reliability and safety certification become problematic due to insufficient evidence of safety standards
Solution Approach 1:
The patent applies preliminary action by performing validation during the training phase before the algorithm is deployed for safety-critical operations. The system validates the trained algorithm against a validation data set and determines whether it meets safety requirements in advance, ensuring safety certification is achieved before operational use.
Solution Approach 2:
The patent implements feedback by continuously monitoring the performance of the machine learning algorithm during training and validation phases. The system provides feedback on whether the algorithm meets the safety threshold, allowing for iterative improvement and adjustment of the model to ensure it satisfies safety requirements.
2Reliability
If the training phase is extended to improve model accuracy, then the reliability of the algorithm increases, but the time and computational resources required increase
Solution Approach 1:
The patent applies preliminary action by integrating validation into the training phase itself, rather than performing it separately after training is complete. This allows the system to determine early whether the trained algorithm meets safety requirements, avoiding unnecessary extension of the training phase and reducing overall time consumption.
3Reliability
If safety validation is performed rigorously to ensure safety standards are met, then the reliability of the system improves, but the complexity of the validation process increases
Solution Approach 1:
The patent merges the validation process with the training phase, combining what were previously separate processes into a unified workflow. The validation data set is prepared in advance, and validation is performed automatically during training, reducing the complexity of managing separate validation processes while maintaining rigorous safety standards.
Data Source
AI summary
A supervised machine learning of a computer-implemented method performs a technical process in which a data pool is created, containing data sets with input data and output data that describes a correct process result. The data sets are divided into training and validation data sets. The method is trained in a training phase, wherein process parameters of the method are varied during repeated performances of the method. A trained method is checked in a validation phase by comparing the output data calculated with the trained method with the input data, with the output data describing the correct process result and calculating an actual probability of failure for the method as a result of the comparison. An empirical probability of failure is defined for the training phase, which is at most as large as a specified target probability of failure.