transparency

An algorithm (in particular machine learning) is said to be transparent if humans can make sense of its internal processes. The opposite to transparency is a blacx-box model, wherethe compexity of the model means one simply has to accepts the outputs on trust. Transparency may be inherent in the kind of model, such as (small) decison trees or rule-based systems, or due to applying techniques for explainable AI to otherwise a blacx-box model. When cosidering machine learning, transparency may apply to:

decision rules -- whether the final outcome of machine learning is comprehensible; for example a small set of rules (as opposed to, say, a vast number of neural network weights)

learning processn -- whether the algorithm used to detrmine the decsion rules is coprehensible; for example a top-down tree-learning algorithm such as ID3 (as opposed to, say, a genetic algorithm).,

Used on pages 476, 482, 491, 507, 508, 509, 559

Also known as transparent