Improving Deep Knowledge Tracing via Gated Architectures and Adaptive Optimization
Altun Shukurlu
TL;DR
The paper addresses improving Deep Knowledge Tracing by integrating gated architectures (LSTM and GRU) and transitioning to a PyTorch implementation to enhance reproducibility. It systematically benchmarks optimization algorithms (SGD, RMSProp, Adagrad, Adam, AdamW) on the DKT task using Synthetic-5 and Khan Academy data, demonstrating that gated models better capture long-range dependencies and that adaptive optimizers, particularly AdamW, accelerate convergence and improve predictive accuracy. The work provides a reusable PyTorch foundation for neural knowledge tracing, with practical implications for real-time personalized learning, where GRU offers speed advantages and AdamW yields the best speed–accuracy trade-off. Future work points to incorporating newer optimizers and attention mechanisms, while improving interpretability of knowledge trajectories.
Abstract
Deep Knowledge Tracing (DKT) models student learning behavior by using Recurrent Neural Networks (RNNs) to predict future performance based on historical interaction data. However, the original implementation relied on standard RNNs in the Lua-based Torch framework, which limited extensibility and reproducibility. In this work, we revisit the DKT model from two perspectives: architectural improvements and optimization efficiency. First, we enhance the model using gated recurrent units, specifically Long Short-Term Memory (LSTM) networks and Gated Recurrent Units (GRU), which better capture long-term dependencies and help mitigate vanishing gradient issues. Second, we re-implement DKT using the PyTorch framework, enabling a modular and accessible infrastructure compatible with modern deep learning workflows. We also benchmark several optimization algorithms SGD, RMSProp, Adagrad, Adam, and AdamW to evaluate their impact on convergence speed and predictive accuracy in educational modeling tasks. Experiments on the Synthetic-5 and Khan Academy datasets show that GRUs and LSTMs achieve higher accuracy and improved training stability compared to basic RNNs, while adaptive optimizers such as Adam and AdamW consistently outperform SGD in both early-stage learning and final model performance. Our open-source PyTorch implementation provides a reproducible and extensible foundation for future research in neural knowledge tracing and personalized learning systems.