Thuật ngữ AI
Từ điển đầy đủ về Trí tuệ nhân tạo
Algorithmic Accountability
Formal mechanism for clearly assigning responsibility for decisions made by an AI system to the various stakeholders involved in its design and deployment.
Algorithmic Debiasing
Set of techniques aimed at identifying, measuring, and correcting systematic biases present in AI models to ensure fairer outcomes.
Distributive Equity
Ethical principle ensuring that the benefits and drawbacks of algorithmic decisions are distributed fairly among different demographic groups.
AI Governance
Organizational and regulatory framework defining responsibilities, processes, and controls to ensure the ethical development and deployment of AI systems.
Algorithmic Impact Assessment
Systematic evaluation of the potential effects of an AI system on fundamental rights, equality, and non-discrimination before its deployment into production.
Algorithmic Procedural Justice
Principle ensuring that automated decision-making processes are transparent, consistent, and allow for meaningful participation of affected individuals.
Redress Mechanism
Formal procedure that allows individuals to challenge an algorithmic decision and obtain a human review or a correction in case of error.
Diffuse Accountability
Ethical problem where the accountability for an automated decision is shared among multiple actors, making it difficult to clearly assign fault.
Operational Transparency
Obligation to document and communicate decision-making processes, data used, and limitations of AI systems to relevant stakeholders.
Formal Verification
Rigorous mathematical approach to prove that algorithms formally comply with specified properties, ensuring their expected behavior.
Algorithmic Uncertainty Zone
Operating domain where the model shows low confidence in its predictions, requiring human intervention or additional precautions.