Four patent applications. Two arXiv preprints. Two papers in preparation. The IP foundation under the structure-centric paradigm is being built deliberately, with priority dates established before public disclosure of implementation details.
| Title | Status | Filing |
|---|---|---|
| AdaBox — structure-centric clustering algorithm and scale-invariant parameter formulation | Filed | January 2026 |
| SCOPE — decomposable evaluation framework for density-based clustering | Filed | January 2026 |
| AdaGraph — native high-dimensional clustering, including SLCD pipeline, Density-Aware Sampler, two-pass prototype deployment, and subspace probe (six independent claims plus 22 dependent) | Filed | May 2026 |
| Graph-SCOPE — unsupervised decomposable validity metric for high-dimensional clustering | Filed | May 2026 |
The AdaGraph filing alone protects six distinct inventive concepts as independent claims: the core AdaGraph algorithm, the scale-invariant parameter formulation, the Density-Aware Sampler, the two-pass prototype deployment method, the subspace probe for automatic dimensionality assessment, and the complete SLCD pipeline. Each is independently licensable.
AdaGraph and the Structure-Centric Machine Learning Paradigm
The canonical reference for the structure-centric paradigm. This paper formally introduces SC-ML as a unified framework, presents AdaGraph — the native high-dimensional clustering algorithm operating in 100–5,000+ dimensional spaces without dimensionality reduction — and establishes the full technical foundations of the paradigm across text, cancer genomics, and materials science. This is the paper to cite when referencing Structure-Centric Machine Learning.
Structure-Centric Density-Based Clustering with Scale-Invariant Parameters
The priority-establishing preprint for the AdaBox algorithm and the SCOPE evaluation framework — the foundational low-dimensional building blocks of the SC-ML stack. Filed before full disclosure to establish IP priority. AdaGraph (arXiv:2605.16320) supersedes this as the primary paradigm reference.
Mathematical formulation of the five-component decomposition; benchmark against Silhouette, Davies-Bouldin, Calinski-Harabasz across 10 synthetic and 5 real-world datasets; dimensionality scaling proof. Target venue: NeurIPS 2027.
The C6 module discovery on GSE31210, literature validation of all 44 genes, and laboratory validation of the highest-priority candidates. Co-authored with cancer genomics laboratory. Target venue: a leading genomics or oncology journal.