Real-Time IoT Pond Water Dataset for Fish Farming

This paper introduces a real-time water quality dataset of five ponds for fish farming obtained through an IoT framework for monitoring the aquatic environmental conditions. It utilizes sensors and an Arduino microcontroller to collect data on pH, temperature, and turbidity in pond water in Jamalpur District, Bangladesh. The data is stored in an IoT cloud platform named ThingSpeak and analyzed using 10 machine learning algorithms. The dataset consists of 4 columns and 40280 rows, where pH, temperature, turbidity, and fish are recorded. Fish represents the target variable, while the others serve as independent variables. Within the dataset, there are 11 distinct fish categories including sing, silver carp, Katla, prawn, karpio, shrimp, rui, pangas, tilapia, magur, and koi. Results showed that only three ponds are suitable for fish farming among five ponds and the Random Forest algorithm performs the best. The study also includes details of the IoT system's hardware. This dataset will be useful for researchers and fish farmers to predict fish survival.