A dual-strategy multi-objective evolutionary algorithm with domain-informed mutation and secondary weighted selection for counterfactual explanations in knowledge tracing

Key Points

• Counterfactual explanations improve understanding in knowledge tracing, aiding educational interventions.
• The primary metric shows enhanced accuracy using a multi-objective evolutionary algorithm with domain-informed mutation.
• This study employs a dual-strategy approach combining weighted selection and informed mutation to optimize algorithm performance.
• Further validation in real-world educational settings is needed to establish broader applicability.