博碩士論文 110225013 完整後設資料紀錄

DC 欄位 語言
DC.contributor統計研究所zh_TW
DC.creator劉彥劭zh_TW
DC.creatorYan-Shao Liouen_US
dc.date.accessioned2023-7-8T07:39:07Z
dc.date.available2023-7-8T07:39:07Z
dc.date.issued2023
dc.identifier.urihttp://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=110225013
dc.contributor.department統計研究所zh_TW
DC.description國立中央大學zh_TW
DC.descriptionNational Central Universityen_US
dc.description.abstract本論文提出了貝氏在線變點檢測方法基於Clayton copula 且邊際分布為混合 常態分布的copula-Markov 模型。在模擬研究中,我們研究了相關時間序列數據 的結構變化,包括單峰數據和雙峰數據之間的轉換以及雙峰數據內部的結構變 化,並將其與邊際分布為常態分布的copula-Markov 模型進行比較,模擬的結果 指出在相關性資料由單峰改變成雙峰時,我們所使用的模型在準確率、平均絕對 誤差、真陽個數、偽陽個數皆勝過對比模型。我們應用這種方法來檢測加密貨幣 市場從2017 年到2018 年初的變點,其中包括ICO 繁榮的開始和結束,我們選用 BTC、ETH、NEO 的每日報酬率進行偵測。結果顯示了我們的方法在檢測相關時 間序列數據的變化方面的有效性,並提供了關於虛擬貨幣市場在這ICO 繁榮-蕭條 期間行為的見解。 關鍵字:在線改變點檢測,Clayton copula,馬可夫模型,貝氏推論,ICO boomzh_TW
dc.description.abstractIn this paper, we propose an online Bayesian changepoint detection approach for dependent time series data under the copula-based Markov model with the marginal distributions being mixture normal distributions. Simulation studies examine structural changes in sequential time series data with dependency, including transformation between unimodal and bimodal data and structural changes within bimodal data. For comparison, we consider the Clayton copula-based Markov model with normal marginal distributions as the benchmark model. The results show that the proposed model outperforms the benchmark model in detecting change when the correlation structure of the changes from unimodal data to bimodal data. In the empirical analysis, we use the daily returns of BTC, ETH, and NEO to identify change points in the cryptocurrency market from 2017 to early 2018. The results demonstrate the effectiveness of our approach in detecting changes and providing insight into cryptocurrency market behavior during ICO booms and busts. Keywords: online change point detection, Clayton copula, Markov model, Bayesian inference, ICO boomen_US
DC.subject在線改變點檢測zh_TW
DC.subjectClayton copulazh_TW
DC.subject馬可夫模型zh_TW
DC.subject貝氏推論zh_TW
DC.subjectICO boomzh_TW
DC.subjectonline change point detectionen_US
DC.subjectClayton copulaen_US
DC.subjectMarkov modelen_US
DC.subjectBayesian inferenceen_US
DC.subjectICO boomen_US
DC.titleOnline Change Point detection under a Copula-Based Markov Chain Model for Bimodal Time Seriesen_US
dc.language.isoen_USen_US
DC.type博碩士論文zh_TW
DC.typethesisen_US
DC.publisherNational Central Universityen_US

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡  - 隱私權政策聲明