Explainable Regime Conditioned Mixture of Experts for Financial Anomaly Detection: Replication Dataset

Derived rolling-window features, engineered factor channels, and the exact values underlying the reported means, standard deviations, tables, and figures in Hemrajani et al., 'Explainable Regime Conditioned Mixture of Experts for Financial Anomaly Detection, Risk Capital Calibration, and Cross Market Validation.' Raw source market data was obtained via the yfinance API; risk-factor data from the Kenneth R. French Data Library. See the included README.md for a full file-by-file description. Version 2 (second revision, Applied Artificial Intelligence Ms 262098302.R1) adds R2_replication_data_v2.zip: daily OHLCV data from 2005 (Bitcoin from 2014), the S&P 500 total-return index, the 13-week T-bill rate and the Fama-French factors used as inputs, and the out-of-sample scores, risk forecasts, backtest series and result files produced by code release v2.0.0. The version 1 files are unchanged. See README.txt inside the zip.