Technological Innovation

What is BS EN 45536-2017?

BS EN 45536-2017 is a technical standard developed by the British Standards Institution (BSI) that provides guidelines for the evaluation and validation of machine learning algorithms. In an increasingly AI-driven world, this standard plays a crucial role in ensuring the reliability and safety of AI systems.

Evaluating Machine Learning Algorithms

The first key aspect covered by BS EN 45536-2017 is the evaluation process of machine learning algorithms. This involves assessing the algorithm's performance and capabilities within specific contexts and data sets. The standard offers guidance on selecting appropriate evaluation metrics and benchmarks to ensure reliable comparisons between different algorithms.

Furthermore, BS EN 45536-2017 emphasizes the importance of transparency and interpretability in algorithmic decision-making. It encourages developers to thoroughly document the decision rules, training procedures, and hyperparameter settings used in developing their algorithms. This documentation enables third-party experts to evaluate the algorithms comprehensively and identify potential biases or ethical concerns.

Validation Processes for AI Systems

The second significant aspect addressed by BS EN 45536-2017 is the validation of AI systems. Validation ensures that the deployed system performs as intended and meets the defined requirements. The standard outlines processes for verifying both functional and non-functional aspects of AI systems, such as accuracy, reliability, robustness, and security.

One important consideration highlighted in BS EN 45536-2017 is the need for interpreting and explaining the reasoning behind AI system outputs. This is especially relevant for applications in critical domains like healthcare or finance. The standard suggests methods for generating explanations that help users understand why a particular decision was made, thus building trust and acceptance of AI-based solutions.

Ensuring Ethical and Responsible AI

BS EN 45536-2017 also addresses the ethical aspects of AI deployment. It emphasizes the need for developers to consider and mitigate potential biases and discrimination in machine learning algorithms. The standard promotes fair treatment, equal opportunities, and non-discriminatory practices in AI system design and implementation.

Additionally, BS EN 45536-2017 advocates for continuous monitoring and management of AI systems post-deployment. This ensures that any issues or shortcomings can be identified promptly and appropriate corrective actions can be taken. Regular audits of the AI system's performance and impact are recommended to uphold ethical standards throughout the system's lifecycle.


BS EN 45536-2017 plays a vital role in promoting transparency, reliability, and ethical responsibility in the development and deployment of AI systems. By providing guidelines for algorithm evaluation, validation processes, and ensuring ethically sound practices, this standard helps build public trust and confidence in AI technologies. Adhering to BS EN 45536-2017 enables developers to create AI solutions that are more accountable, explainable, and aligned with societal values.


Contact: Cindy

Phone: +86-13751010017


Add: 1F Junfeng Building, Gongle, Xixiang, Baoan District, Shenzhen, Guangdong, China

Scan the qr codeclose
the qr code
TAGS Test Probe BTest Probe 18Test Probe 11Go GaugesIEC 61032IEC 60335Test PinTest FingerIEC 60061-3Wedge Probe7006-29L-47006-27D-37006-11-87006-51-27006-51A-2 7006-50-17006-27C-17006-28A-1Test Probe7006-27B-1IEC 61010IEC 60529IEC 60068-2-75