Bias

Algorithmic bias refers to systematic and unfair discrimination in algorithmic decision-making processes.

Anchoring Bias in AI refers to the cognitive tendency to rely heavily on the first piece of information encountered.

Exposure bias refers to the tendency of algorithms to favor overrepresented data in training sets, affecting model performance.

Implicit Bias Amplification refers to the unintended reinforcement of existing biases in AI systems.

Label bias refers to the systematic errors in labeling data that can affect AI model performance.

Learning bias refers to systematic errors in AI models due to skewed training data or design choices.

Measuring bias involves assessing the fairness and impartiality of AI systems in decision-making processes.

Model bias occurs when an AI model produces systematic errors due to skewed training data or flawed assumptions.

Overestimation Bias is the tendency to overrate one's abilities, knowledge, or predictions.

An overrepresented class in AI refers to a category that appears more frequently in data than others, impacting model bias.