Results for "risk"
Empirical Risk Minimization
IntermediateMinimizing average loss on training data; can overfit when data is limited or biased.
Empirical Risk Minimization is like trying to get the best score on a test by practicing with sample questions. When you practice, you want to get as many answers right as possible, which is similar to how a machine learning model learns from its training data. However, if you only focus on those...
Quantifying financial risk.
Central log of AI-related risks.
Risk of incorrect financial models.
Grouping patients by predicted outcomes.
AI used in sensitive domains requiring compliance.
Existential risk from AI systems.
European regulation classifying AI systems by risk.
US framework for AI risk governance.
Classifying models by impact level.
Minimizing average loss on training data; can overfit when data is limited or biased.
Categorizing AI applications by impact and regulatory risk.
Maximum expected loss under normal conditions.
Risk threatening humanity’s survival.
Framework for identifying, measuring, and mitigating model risks.
Required human review for high-risk decisions.
International AI risk standard.
Models estimating recidivism risk.
Simulating adverse scenarios.
Predicting borrower default risk.
Privacy risk analysis under GDPR-like laws.
Restricting distribution of powerful models.
A function measuring prediction error (and sometimes calibration), guiding gradient-based optimization.
How well a model performs on new data drawn from the same (or similar) distribution as training.
Randomly zeroing activations during training to reduce co-adaptation and overfitting.
Breaking documents into pieces for retrieval; chunk size/overlap strongly affect RAG quality.
Training across many devices/silos without centralizing raw data; aggregates updates, not data.
Systematic review of model/data processes to ensure performance, fairness, security, and policy compliance.
Samples from the k highest-probability tokens to limit unlikely outputs.
Constraining model outputs into a schema used to call external APIs/tools safely and deterministically.
A narrow minimum often associated with poorer generalization.