Results for "biases"

Structured dataset documentation covering collection, composition, recommended uses, biases, and maintenance.

A parameterized mapping from inputs to outputs; includes architecture + learned parameters.

The learned numeric values of a model adjusted during training to minimize a loss function.

Minimizing average loss on training data; can overfit when data is limited or biased.

Standardized documentation describing intended use, performance, limitations, data, and ethical considerations.

Logging hyperparameters, code versions, data snapshots, and results to reproduce and compare experiments.

Ensuring decisions can be explained and traced.

Simplified Boltzmann Machine with bipartite structure.

Required descriptions of model behavior and limits.

Patient agreement to AI-assisted care.

Requirement to provide explanations.

Models estimating recidivism risk.