What is Explainable AI?
Explainable artificial intelligence (XAI) comprises processes and methods enabling users to comprehend and trust machine learning algorithm outputs. It describes AI models, their expected impact, and potential biases while characterizing model accuracy, fairness, transparency, and decision-making outcomes.
Why it matters
XAI is essential for organizations deploying AI models into production, building trust and confidence in those systems. Understanding how an AI system produces specific outputs lets developers verify proper functioning, organizations meet regulatory requirements, and affected parties challenge or modify decisions.
See how Diwo operationalizes Explainable AI.
Read the decision-intelligence playbooks that put this concept to work at Fortune 50 scale.
Related concepts
Artificial intelligence (AI) is the simulation of human intelligence processes by machines, especially computer systems. AI employs sophisticated analysis and logic-based techniques—including machine learning—to interpret events, facilitate and automate decisions, and execute actions. This technology enables machines to understand, respond to, and learn with human-comparable levels of intelligence.
Machine learning represents a data analytics approach that harnesses artificial intelligence to emulate human learning from experience. Rather than depending on preset formulas, machine learning algorithms extract "knowledge" directly from data through computational methods. These algorithms identify inherent patterns within datasets that generate actionable insights and support improved forecasting and decision-making.
Evidence represents the pertinent data and analytical logic enabling users to authenticate recommendations from a decision intelligence system. It encompasses the insights that prompted the recommendation alongside supporting data utilized by models and the DI platform. Evidence frequently aligns with essential metrics and KPIs, incorporating projected impacts of recommendations expressed using the same measurement terms.
Data analytics is the systemic computational analysis of data. It is used for the discovery, interpretation and communication of meaningful patterns in data. The discipline also encompasses applying discovered data patterns to inform effective business decision-making.