Unveiling School Safety: Insights from COVID-19 and School Datasets in Surabaya City, Indonesia

Description

Our dataset comprises two distinct sets of information: one detailing the distribution of COVID-19 cases across 154 sub-areas in Surabaya City, Indonesia, spanning from March 2020 to the end of October 2022, and the other containing school data for each sub-area. The COVID-19 dataset, originally consisting of 146,299 data rows, underwent a pre-processing phase to eliminate erroneous data, resulting in a clean dataset with 146,288 data rows. Simultaneously, the school dataset encompasses information from 1,466 schools. The processing of these datasets is instrumental in gauging the COVID-19 case ratio in each sub-area, facilitating accurate predictions for school openings during the pandemic. This dataset is subjected to several algorithms employing deep learning based on an artificial intelligence approach, with 70% of the data allocated for training and 30% for data testing. Moreover, by engaging in hyperparameter tuning, the prediction model based on a deep neural network is responsible for examining the data and determining appropriate decisions. Hyperparameter tuning generates an optimal prediction model that minimizes the loss function on independent data, mainly when dealing with extensive datasets. Subsequently, the information about predicted COVID-19 spread is presented in a dashboard system, streamlining monitoring and serving as a tool for schools to determine the optimal time for reopening—emphasizing a minimum case ratio, indicative of low virus transmission and safety. Our extensive dataset, optimized with deep learning algorithms and real-time dashboard visualization, can benefit various stakeholders beyond education in future severe outbreaks.

Institutions

Categories

Funding

Licence