25 structural break detection methods for univariate time series: XGBoost, Neural Networks, Ensembles, Reinforcement Learning, and Statistical approaches. Evaluated on cross-dataset generalization.
Fundamental package for quantitative finance with Python.
Official Python implementation of the Pioneer Detection Method (PDM) — convergence-based expert aggregation and opinion pooling under structural change. Code for Vansteenberghe (2026), The Geneva Papers 51(1).
CUSUM (Cumulative Sum) filter for detecting structural shifts in financial time series, implemented in Python
Long-run analysis (1975-present) of public and private investment in Portugal using European Commission AMECO data: robust DBnomics pipeline, structural-break and lagged-regression analysis, real (chain-linked) series, and cross-country comparison (Spain, Greece, Ireland, EU).
Final model for the ADIA Lab Structural Break Challenge 2026: 23rd of 729 teams, solo, 66.01 points where guessing scores 50. Code, weights, tests and write-up.
Bai–Perron structural break detection and estimation for time series and panel data. Tests for breaks, estimates break dates with confidence intervals, and selects break counts via sequential testing or information criteria.
R package for testing and estimating structural breaks in time series and panel data using Bai-Perron and dynamic programming methods. Provides hypothesis tests (supF, UDmax, WDmax), break-date estimation, and support for fixed effects and common correlated effects models.
Official vs. interpolated quarterly data in Timor-Leste: a Bayesian VAR and local-projection study of fiscal transmission and import leakage. Includes temporal disaggregation, structural breaks, IRFs/FEVDs, and a reproducible R/Shiny dashboard.
Credit risk forecasting using unemployment, cointegration, ECM, structural breaks and out-of-time model validation. Chilean banking data.
Loss cost trend analysis for insurance pricing — frequency/severity decomposition, ONS index integration, structural break detection (154 tests)
End-to-End Python implementation of Mukhia et al.'s (2025) methodology for detecting political risk transmission in stablecoin markets. Implements dynamic programming for endogenous breakpoint detection, Empirical Mode Decomposition, Cholesky-identified structural shocks, and AAFT surrogate validation to quantify political uncertainty spillovers.
My experiments for the ADIA Lab Structural Break Challenge (CrunchDAO, 2025): detecting structural breaks in univariate time series with statistical-test features, gradient boosting, and deep learning (Siamese, CNN-BiLSTM, attention).
Detecting regime changes in financial time series using Chow, CUSUM, and Bai-Perron. Break-aware forecasting with ARIMA, Prophet, and LSTM compared against naive baselines.
A Python-based analysis of fundamental econometric challenges, including Omitted Variable Bias (OVB), sensitivity to outliers, model selection criteria, elasticity, and time series stationarity, utilizing both simulated environments and empirical real-world financial data.
Daily Svensson estimation of the French OAT zero-coupon curve, 1987-2026: political risk, PCA by regime and regime detection. MSc Finance thesis, University of Pavia
Unit root tests (ADF, KPSS, Zivot–Andrews) for Turkish monthly card spending, USD loan rates and prices, 2014–2024
Real-time structural-break detection in time series: whitening by the series' own history, CUSUM/GLR/Shiryaev-Roberts on the whitened stream, LightGBM rankers and causal TCNs. ~Rank 160 of 1,716 in the ADIA Lab / CrunchDAO challenge, with the full 158-experiment research record run with an LLM assistant.
R time series analysis of Google Books Ngram usage of "peace" using breakpoints, Markov switching, and ARIMA.
To associate your repository with the structural-breaks topic, visit your repo's landing page and select "manage topics."