Learn more about philosophy with this collection
Understanding machine learning models
Improving data analysis and decision-making
How Google uses logic in machine learning
Training-serving skew
The difference between performance during training and performance during serving—is a persistent challenge.
During training, try to identify potential skews and work to address them, including by adjusting your training data or objective function. During the evaluation, continue to try to get evaluation data that is as representative as possible of the deployed setting.
30
255 reads
MORE IDEAS ON THIS
Directly examine your raw data
ML models will reflect the data they are trained on, so analyze your raw data carefully to ensure you understand it.
- Does your data contain any mistakes (e.g., missing values, incorrect labels)?
- Is your data sampled in a way that represents your users and real-world settings?
32
278 reads
Use a human-centred design approach
The way actual users experience your system is essential to assessing the true impact of its predictions, recommendations, and decisions.
- Design features with appropriate disclosures built-in: clarity and control are crucial to a good user experience.
- Consider augmentation and ...
28
354 reads
Identify multiple metrics to assess training and monitoring
The use of several metrics rather than a single one will help you to understand the tradeoffs between different kinds of errors and experiences.
Consider metrics including feedback from user surveys, quantities that track overall system performance and short- and long-term product health (e...
27
303 reads
Test, test, test
- Conduct rigorous unit tests to test each component of the system in isolation.
- Conduct integration tests to understand how individual ML components interact with other parts of the overall system.
- Proactively detect input drift by testing the statistics of the inputs to th...
32
236 reads
Responsible AI practices
The development of AI is creating new opportunities to improve the lives of people around the world, from business to healthcare to education.
It is also raising new questions about the best way to build fairness, interpretability, privacy, and security into these systems.
29
456 reads
Monitor and update the system after deployment
- Issues will occur: any model of the world is imperfect almost by definition. Build time into your product roadmap to allow you to address issues.
- Consider both short- and long-term solutions to issues. A simple fix (e.g., blocklisting) may help to solve a problem quickly, but may not...
28
251 reads
Understand the limitations of your dataset and model
- A model trained to detect correlations should not be used to make causal inferences, or imply that it can.
- Machine learning models today are largely a reflection of the patterns of their training data.
- Communicate limitations to users where possible.
30
252 reads
General recommended practices for AI
Reliable, effective user-centered AI systems should be designed following general best practices for software systems, together with practices that address considerations unique to machine learning.
27
466 reads
CURATED FROM
CURATED BY
Read & Learn
20x Faster
without
deepstash
with
deepstash
with
deepstash
Access to 200,000+ ideas
—
Access to the mobile app
—
Unlimited idea saving & library
—
—
Unlimited history
—
—
Unlimited listening to ideas
—
—
Downloading & offline access
—
—
Personalized recommendations
—
—
Supercharge your mind with one idea per day
Enter your email and spend 1 minute every day to learn something new.
I agree to receive email updates