Machine learning assurance

Machine learning assurance from DNV GL

Machine learning assurance provides a risk assessment approach to establish trust in systems that use machine learning (ML).

Machine learning assurance

Interest in systems incorporating machine learning, artificial intelligence (AI) and other data-driven techniques is increasing. During the development of these systems, it is challenging to always make good decisions and choices (of algorithms, techniques, data, evaluation criteria etc.). The wrong choices will lead to poor results. As a stakeholder, you want evidence that the system is going to work as expected and will meet the business needs of the users. How can you obtain this evidence and achieve machine learning assurance?

DNV GL's machine learning assurance approach

DNV GL has developed an approach to machine learning assurance that consists of the following:

  1. Assess the development process
  2. Assess the model
  3. Assess the risk of using the model as intended

This approach, based on our forthcoming Recommended Practice DNVGL-RP-0510 Framework for assurance of data-driven algorithms and models, can help stakeholders in machine learning projects identify and manage risk and maximize the likelihood of a successful outcome.

Using a workshop and questionnaire-driven format, the service delivers:

  • A detailed register of risk items and mitigating actions
  • Feedback to your machine learning project team or data science vendor from DNV GL’s domain experts and data scientists
  • Ultimately, trust and assurance that the machine learning model will deliver the expected results

This methodology can help you compare machine learning projects and achieve machine learning assurance of your systems.

Understanding risk of machine learning projects

We help you with a framework that eases communication about a machine learning project’s risk to all stakeholders – it provides easy-to-understand information about complex data science topics. The machine learning assurance and risk assessment covers the complete pipeline, from data collection and ingest to data preparation, modelling, prediction and deployment. For each stage in the pipeline, aspects of high risk are identified and mitigating actions suggested.

Contact us:

How can we help you?

Request information

Know what you need?

Request quote

How can we help you with data analytics?

Contact us

Related services you might find interesting: