ANALYZING TEMPEARTURE ANOMALIES IN MONITORING DATA USING CONVOLUTIONAL NEURAL NETWORK

Temperature is a tool that shows the degree or measure of how hot or cold an object is. Incorrect temperature measurement can be fatal and cause various problems. Abnormal temperatures can prevent the temperature detection system from running optimally. Therefore, it is necessary to classify temperatures into normal and anomalous. Machine learning can be used as an alternative for temperature classification. By utilizing machine learning methods, one of which is Convolutional Neural Network. 3688569 temperature data were tested, dividing the results into 80% training data and 20% testing data. Accuracy, Precision, Recall, and F1 Score get a score of 100% and the CNN model graph is very good.