How to forecast daily carbon emissions during public health emergencies: A novel self-attention multi-neuron time series model

Abstract

Affected by numerous uncertainties, climate change is a critical issue linked to carbon emissions that warm the planet. Although scholars have conducted detailed research on carbon emissions and established predictive models for them, there are few models specifically designed for predicting carbon emissions during public health emergencies. With the concentrated outbreak of various uncertain factors, organizations and institutions urgently need a model capable of predicting carbon emissions during public health emergencies. This study introduces a novel self-attention multi-neuron time series (SAMNTS) model to evaluate the previously unexplored impact of public health emergencies on carbon emissions. Specifically, we have designed a more comprehensive deep learning prediction framework that can effectively utilize a large amount of relevant data to conduct detailed reasoning and analysis on the issue of carbon emissions, enabling more accurate predictions of daily carbon emissions. To better test its effectiveness, we used COVID-19 as an example to test the model. The results proved that the model can effectively make predictions and analyze various factors that affect carbon emissions.