The Data-Centric AI Approach With Andrew Ng

The presentation discusses the development of data-centric AI and provides tips for its implementation, with a focus on unstructured data.

• Data-centric AI is becoming more widespread and systematic in its approach
• Consistent labeling of data is crucial for learning algorithms to work effectively
• Error analysis and engineering examples are important for structured data
• Data augmentation and noise examples can be useful for unstructured data
• Focusing on subsets of data can improve performance

The speaker discusses the importance of consistent labeling of data for learning algorithms to work effectively, using the example of visual defect inspection in manufacturing. Inconsistent labeling can lead to inaccurate results, even among human expert inspectors. By re-labeling images based on scratch length, the data set can be made more consistent and accurate.

Dr. Andrew Ng, Founder of DeepLearning.AI and Founder and CEO of Landing AI shares real-world examples to show how switching from a model-centric to a data-centric AI development approach, by 'engineering the data', helps improve the performance of machine learning models. He shares his views on key AI best practices for those considering data-centric model development to unlock greater AI performance and efficiencies.