Engineering performant and trustworthy AI solutions : ensuring AI product quality and reliability.

Format:

Video

Contributor:

Mohanna, Ammar, instructor.

O'Reilly (Firm), publisher.

Language:

English

Subjects (All):

Artificial intelligence--Quality control.

Artificial intelligence.

Machine learning--Evaluation.

Machine learning.

Artificial intelligence--Testing.

Physical Description:

1 online resource (1 video file (2 hr., 27 min.)) : sound, color.

Edition:

[First edition].

Place of Publication:

[Sebastopol, California] : O'Reilly Media, Inc., [2024]

Summary:

This course focuses on the unique challenges and methodologies involved in testing and validating AI and machine learning models. It provides a comprehensive understanding of the paradigms and practices essential for assuring the quality and reliability of AI-powered products. The course covers the technical, practical, and business perspectives of AI QA, offering participants the tools and knowledge needed to enhance their AI development processes. As AI technologies become increasingly integral to various industries, ensuring their reliability and performance is crucial. Quality assurance in AI is not just about verifying accuracy but also about addressing issues like data quality, algorithmic bias, and model explainability. For AI developers, engineers, and QA professionals, mastering these aspects is vital to delivering robust, market-ready AI solutions that meet business objectives and user expectations. This course addresses the specific challenges of testing AI systems, including handling non-deterministic outputs, managing data biases, and ensuring continuous learning and adaptation. It provides practical solutions for integrating QA processes into the AI development lifecycle, helping professionals mitigate risks, enhance model performance, and maintain ongoing reliability. By understanding and applying effective QA strategies, participants can overcome common obstacles in AI projects, ultimately leading to more successful deployments.

Notes:

OCLC-licensed vendor bibliographic record.

OCLC:

1463911228

Publisher Number:

0642572060008