A Systematic Review of CNN-LSTM-Family Methods, Datasets and Transferability : Supplementary
Published: 9 September 2026| Version 1 | DOI: 10.17632/g8ytpvksw6.1
Contributor:
Karan ParmarDescription
Supplementary data for the systematic review "Deep Learning for Flood Prediction: A Systematic Review of CNN-LSTM-Family Methods, Datasets and Transferability (2015–2026)," reported per PRISMA 2020. A single Excel workbook contains four sheets: the five Boolean search strings and their per-database record counts (S1), the keyword relevance-scoring scheme used as an AI-assisted screening aid (S2), the full extraction matrix for all 141 included studies (S3), and the seven-item risk-of-bias ratings for each study (S4). Together these reproduce every count and figure reported in the review. Publisher abstracts have been removed to avoid redistributing copyrighted text.
Files
Categories
Artificial Intelligence, Hydrology, Flood, Weather