Early student grade prediction in education is crucial for timely intervention. AcaPerforma is developed using a novel MSECosine-based Long Short Term Memory model with Generative Adversarial Network which utilises various feature combinations comprising of demography, learning traces, and assessment. The interactive machine learning approach ensures model explainability, transparency and bias mitigation to ensure responsible artificial intelligence governance.

AcaPerforma is a continuation of works on AI in education at UPM since 2018 and have won many awards and protected by copyrights. AcaPerforma could give early course grade prediction which can support for personalised academic intervention