The COVID-19 pandemic that happens worldwide has affected not only human health, social activities, the economy, education but also the death rate caused by this pandemic. Although the death rate from COVID-19 worldwide is quite high, the recovery rate is also quite promising. Therefore, this study is conducted to predict the recovery rate of COVID-19 cases in Indonesia, specifically in Kabupaten Bandung Barat, which was analyzed using the Neural Network Algorithm. The method of this study is data mining, using the neural network algorithm that analyzed data, consisting of 2 attributes and 1 class attribute, namely: Daily Case that represent the daily new confirmed case in the observed location, Daily Death that represents the daily new number of confirmed deaths in observed location. The class attributes are using Daily Recovered, which represents the daily new number of confirmed recoveries in the observed location. The findings of this study indicate that the neural network models in this study have a Root Mean Square Error (RMSE) 102.168 to predict the recovery rate of COVID-19 cases in the observed location.