Action Recommendation Model Development for Hydromon Application Using Deep Neural Network (DNN) Method

Controlling hydroponic plants, which is currently being carried outmanually, can be said to be less effective because it still involves thehard work of farmers to continuously monitor the condition of thehydroponic plants. Therefore, the general objective of this research isto develop a model that can be used as a recommendation system foractions that farmers need to take based on hydroponic crop conditions.The model formed with this machine learning method will then beused in the Hydromon application which allows farmers to manageand monitor the condition of hydroponic plants and take action basedon the recommendations given. This model was developed using adeep neural network algorithm consisting of five layers with the helpof the TensorFlow framework. The results show that the model isaccurate with an accuracy value of 96.47% on the test data to classifyplant conditions so that it can be used in the Hydromon application.