BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//VNP | Chương trình Việt Nam - Hà Lan - ECPv6.17.5//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:VNP | Chương trình Việt Nam - Hà Lan
X-ORIGINAL-URL:https://vi.vnp.edu.vn
X-WR-CALDESC:Events for VNP | Chương trình Việt Nam - Hà Lan
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Asia/Krasnoyarsk
BEGIN:STANDARD
TZOFFSETFROM:+0700
TZOFFSETTO:+0700
TZNAME:+07
DTSTART:20250101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Asia/Krasnoyarsk:20260930T150000
DTEND;TZID=Asia/Krasnoyarsk:20260930T160000
DTSTAMP:20260923T080235Z
CREATED:20260923T074345Z
LAST-MODIFIED:20260923T080235Z
UID:23427-1790780400-1790784000@vi.vnp.edu.vn
SUMMARY:Thesis Public Defense | VNP29 - Võ Thị Kim Nguyệt
DESCRIPTION:Shapelet Discovery in EU Carbon Prices: A Policy Lens for Vietnam \nStudent: Võ Thị Kim Nguyệt\, VNP-30 \nSupervisor: Dr. Trần Thị Tuấn Anh \nAbstract: \nThis study develops a shapelet-based framework for regime detection in the European Union Emissions Trading System (EU ETS). Analysis of 5\,217 daily EU Allowance (EUA) prices from 2005 to 2025 reveals nine distinct market regimes characterized by unique temporal patterns. The proposed methodology achieves 99.68% classification accuracy with an Adjusted Rand Index of 0.850\, substantially outperforming benchmark methods including Gaussian Mixture Model (GMM)\, K-Means\, DBSCAN\, Hidden Markov Models (HMM)\, ARIMA\, and Pruned Exact Linear Time (PELT). The approach identifies 104 discriminative shapelets across multiple temporal scales\, of which 53 achieve statistical significance through Kolmogorov-Smirnov testing with Bonferroni\ncorrection\, ensuring patterns represent genuine regime characteristics rather than spurious correlations.\n \nAnalysis reveals that discriminative patterns occur on average 18.3 days before external regime boundary labels in the historical dataset\, with regime-specific temporal offsets ranging from 12 to 28 days. Nine regimes are characterized: Phase I Formation (2005-2007)\, Phase II Stability (2008-2011)\, Phase III Launch (2013-2014)\, Backloading Anticipation (2014-2015)\, MSR Announcement (2015-2017)\, MSR Active (2018-2019)\, COVID-19 Shock (2020-2021)\, Post-COVID Recovery (2021-2023)\, and Post-COVID Volatility (2023-2025). A conceptual four-phase policy framework for\nVietnam’s carbon market development (2025-2035) is presented based on these findings.\n \nHowever\, all results are derived from in-sample classification without temporal traintest splits. Shapelets were discovered using the complete 2005-2025 dataset with full knowledge of all regime periods. The reported accuracy and temporal patterns represent retrospective classification capability\, not prospective forecasting performance. Out-ofsample validation through walk-forward testing is required before operational deployment. The Vietnam framework represents a conceptual application contingent on successful validation. \nKeywords: Carbon market\, EU ETS\, Shapelet analysis\, Time-series classification\, Regime detection\, Vietnam policy
URL:https://vi.vnp.edu.vn/event/thesis-public-defense-vnp29-v-th-kim-nguyet/
LOCATION:H.001\, 1A Hoàng Diệu\, Thành phố Hồ Chí Minh\, Viet Nam
CATEGORIES:THESIS PUBLIC DEFENSE
ATTACH;FMTTYPE=image/jpeg:https://vi.vnp.edu.vn/wp-content/uploads/2026/09/carbon-price.jpg
END:VEVENT
END:VCALENDAR