BeTiSe Dataset: A Benchmark for Time Series Stationarity and Structural Analysis

Creative Commons License

Yazıcı P. C., Güzel İ., Karagöz P., Yozgatlıgil C., Erkaya P.

  • Yayınlanma Tarihi: 21 Ekim 2026
  • Versiyon: 4
  • Üniversite / Kurum: Orta Doğu Teknik Üniversitesi
  • Fon Sağlayan Organizasyon: ODTU Gap, TUBITAK 1001
  • Araştırma Alanı: İstatistik , Temel Bilimler

Açıklama

BeTiSe is a large-scale synthetic benchmark dataset comprising 120,000 unique univariate time series, specifically designed to address the limitations of traditional statistical tests and the evaluation gaps in modern deep learning architectures, including Time Series Foundation Models and Large Language Models (LLMs).

BeTiSe is a large-scale synthetic benchmark dataset comprising 120,000 unique univariate time series, specifically designed to address the limitations of traditional statistical tests and the evaluation gaps in modern deep learning architectures, including Time Series Foundation Models and Large Language Models (LLMs).

Yeniden Üretme Adımları

Paylaş