基于深度学习的高温环境核心体温预测模型建立与分析
Establishment and analysis of core body temperature prediction model in high temperature environment based on deep learning
摘要：
现阶段核心体温预测大部分都是在常温或是低温环境下,少有高温环境下的核心体温预测方法。随着人工智能的发展,神经网络的出现为温度预测提供了新的技术支持。本文以10名健康男性的体温实测数据为基础,提出了一种基于长短期记忆(LSTM)神经网络的高温环境核心体温预测模型,并与循环神经网络(RNN)模型进行了对比实验。结果表明:与RNN模型相比,LSTM模型具有更高的温度预测准确性,在3种环境温度下的预测精度分别为0.944、0.930、0.913,而RNN模型在3种环境温度下的预测精度分别为0.898、0.865、0.853。
Abstract：
At present, most core body temperature predictions are made in normal or low temperature environment, with few methods for predicting core body temperature in high temperature environment. With the development of artificial intelligence, the emergence of neural networks provides new technological support for temperature prediction. This paper proposes a core body temperature prediction model for high temperature environment based on long short-term memory (LSTM) neural network using measured body temperature data from ten healthy men, and compares it with a recurrent neural network (RNN) model. The results show that compared with the RNN model, the LSTM model has higher temperature prediction accuracy, with prediction accuracy of 0.944, 0.930 and 0.913 at three environment temperatures, respectively, while the prediction accuracy of the RNN model is 0.898, 0.865 and 0.853 at three environment temperatures, respectively.
Keywords:deep learning; high temperature environment; core body temperature; long short-term memory; recurrent neural network