A Graph Tsetlin Machine with hypervectorized macro multigraphs and message-passing clauses anticipates four USD/JPY regimes, reaching 70.7% overall OOS accuracy and beating reduced-graph and CoTM baselines on stagnant and choppy classes.
hub
Explicit expanders of every degree and size
1 Pith paper cite this work, alongside 5,369 external citations. Polarity classification is still indexing.
abstract
An $(n,d,\lambda)$-graph is a $d$ regular graph on $n$ vertices in which the absolute value of any nontrivial eigenvalue is at most $\lambda$. For any constant $d \geq 3$, $\epsilon>0$ and all sufficiently large $n$ we show that there is a deterministic poly(n) time algorithm that outputs an $(n,d, \lambda)$-graph (on exactly $n$ vertices) with $\lambda \leq 2 \sqrt{d-1}+\epsilon$. For any $d=p+2$ with $p \equiv 1 \bmod 4$ prime and all sufficiently large $n$, we describe a strongly explicit construction of an $(n,d, \lambda)$-graph (on exactly $n$ vertices) with $\lambda \leq \sqrt {2(d-1)} + \sqrt{d-2} +o(1) (< (1+\sqrt 2) \sqrt {d-1}+o(1))$, with the $o(1)$ term tending to $0$ as $n$ tends to infinity. For every $\epsilon >0$, $d>d_0(\epsilon)$ and $n>n_0(d,\epsilon)$ we present a strongly explicit construction of an $(m,d,\lambda)$-graph with $\lambda < (2+\epsilon) \sqrt d$ and $m=n+o(n)$. All constructions are obtained by starting with known ones of Ramanujan or nearly Ramanujan graphs, modifying or packing them in an appropriate way. The spectral analysis relies on the delocalization of eigenvectors of regular graphs in cycle-free neighborhoods.
hub tools
fields
cs.CE 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Macroeconomic Message Passing for Anticipating Foreign Exchange Regime Changes: A Deep Logical Learning Approach using Graph Tsetlin Machines
A Graph Tsetlin Machine with hypervectorized macro multigraphs and message-passing clauses anticipates four USD/JPY regimes, reaching 70.7% overall OOS accuracy and beating reduced-graph and CoTM baselines on stagnant and choppy classes.