Results

Each notebook visualises the results produced by the corresponding study script. See the Studies section for the full source code of each script.

• Study A — Impact of Dataset Sparsity \(p_e\) on Convergence
• Study B — Optimal Dataset Sparsity \(p_e^*\) vs Problem Size \(N\)
• Study C — Impact of Learning Rate \(\alpha\) on Convergence
• Study D — Optimal Learning Rate \(\alpha^*\) vs Problem Size \(N\)
• Study E — Impact of \(p_e\) on Convergence (Large Batch)
• Study F — Weight Distribution Dynamics During Training
• Study G — Joint Impact of \(\alpha\), \(p_e\), and Batch Size on Convergence
• Study H — Product Node vs MLP: Sparse vs Full-Table Training