Research on temperature gradient has been carried out in Blawan geothermal area. This study aims to predict the temperature in the subsurface temperature measurement using a temperature probe with a depth of 2 m in the Blawan geothermal area. Temperature and depth are the two variables being measured. Meanwhile, the resistivity, conductivity, and humidity data were taken from previous studies in the exact area measurements. The prediction determination used modeling with an Artificial Neural Network (ANN) with the back-propagation method. The optimal predictions using an Artificial Neural Network (ANN) were obtained by constructing three input layers, five hidden layers, and two output layers (3-5-2) with a hyperbolic tangent function. Results for temperature prediction with the larger R² (1) values and lower MAPE (1.07%), RMSE (0.78), MSE (0.61), and MAD (0.34) values. Moreover, humidity generates a greater R² (1) values and lower MAPE (0.34%), RMSE (0.34), MSE (0.18), and MAD (0.29) values. ANN proved very effective in predicting temperature and humidity factors.

Blawan地热区开展了温度梯度研究。本研究旨在使用 Blawan 地热区 2 m 深度的温度探头预测地下温度测量中的温度。温度和深度是被测量的两个变量。同时，电阻率、电导率和湿度数据取自先前在精确面积测量中的研究。预测确定使用具有反向传播方法的人工神经网络 (ANN) 建模。通过使用双曲正切函数构建三个输入层、五个隐藏层和两个输出层 (3-5-2)，获得了使用人工神经网络 (ANN) 的最佳预测。^{具有较大 R 2}的温度预测结果(1) 值和更低的 MAPE (1.07%)、RMSE (0.78)、MSE (0.61) 和 MAD (0.34) 值。此外，湿度会产生较大的 R ² (1) 值和较低的 MAPE (0.34%)、RMSE (0.34)、MSE (0.18) 和 MAD (0.29) 值。ANN 被证明在预测温度和湿度因素方面非常有效。