Regime Switching

stochastic-process · 37 views
stochastic-processstochastic

What It Is

Two-state hidden Markov process --- alternates between calm (σ=0.5) and volatile (σ=3.0) regimes with asymmetric transition rates

Interpretation

Standard analysis sees: left-skewed; smooth / autocorrelated; long-range memory (persistent); volatility-clustering (bursty); low-dimensional; nonstationary / drifting. The atlas additionally detects Nonstationarity:adf_pvalue. It sits beside ETH/BTC Ratio in the atlas (standard-bank rank 26) — a neighbor conventional features miss.

What standard analysis sees
tail heaviness0.48
asymmetry0.04
occupancy0.61
short-range corr0.94
long-range memory0.93
spectral colour0.28
periodicity0.85
complexity0.69
time-irreversibility0.73
volatility clustering0.95
multifractality0.65
dimensionality0.12
nonstationarity0.93
What the atlas adds
Nonstationarity:adf_pvalue+3.6z
unit-root nonstationarity (ADF cannot reject random-walk null)

Composition

dtypefloat64
range[-45.49, 209.8]
unique values16384 / 16384
mean ± std74.3 ± 54.5

Render Gallery

Atlas Position

Nearest neighborDistance
ETH/BTC Ratio2.39cross-origin
Brownian Walk2.58
Levy Flight3.00

Open in Atlas →

Which Geometries Light Up

Catch24Catch24:PD_PeriodicityWang_th0_01rank 4/3073475.2000
in algorithmic-bytes
alphabetical
← / → within domain · ⇧← / ⇧→ alphabetical · ⇧← / ⇧→ inside an open render = same view across sources