The experimental data contains four parts. The instructions are as follows:

Original data.csv --  Raw data collected through the Internet of things system.
Pretreated data.csv -- The data obtained after preprocessing the original data.
Train dataset.csv -- Data for deep learning model training.
Test dataset.csv -- Test data used to test the performance of a trained model.

If you use these training and test data or find our work valuable, please cite our paper: An Accurate Prediction of Water Quality in Smart Mariculture with Improved SRU Deep Learning Network.Thanks.