超越相关性的隐性市场状态探测（实证结果）
1 分•作者: johnoliveiradev•4 天前
分享一项基于结构耦合而非相关性的实证性状态检测研究。<p>关键发现：相关性可能保持高位，而结构依赖性却会崩溃。
该信号在不同尺度上检测到刚性、过渡和混沌状态。<p>黑盒引擎，仅提供结果。
不具备预测性，未使用机器学习。<p> 欢迎对解读和相关文献提出反馈意见。<p>代码库：https://github.com/johnoliveiradev/Multiscale-structural-regime-benchmark/tree/main/results/BTC%20vs%20ETH%2060%20days
查看原文
Sharing an empirical regime-detection study based on structural coupling rather than correlation.<p>Key finding: correlation can stay high while structural dependence collapses.
The signal detects rigid, transitional and chaotic regimes across scales.<p>Black-box engine, results only.
Not predictive, no ML.<p>Looking for feedback on interpretation and related literature.<p>Repo: https://github.com/johnoliveiradev/Multiscale-structural-regime-benchmark/tree/main/results/BTC%20vs%20ETH%2060%20days