Comparison of Machine Learning Algorithms for Classification of Ultraviolet Index

Indonesia is a tropical climate country which has the potential for high intensity of sunlight exposure. Several types exposure that receives are ultraviolet rays. According to National Agency Meteorology, Climatology and Geophysics, it provides information regarding impact on human activities in an index risk scale. The aim this research was creating recommender system based category providing daily activity advice users. methods used K-nearest Neighbor algorithm Support Vector Machine with Collaborative Filtering Model-Based approach could recommend items results model trained identify input data patterns. stages carried out study included collection, pre-processing, division into test train data, dataset testing, analysis models had been accuracy values using tested. confusion matrix calculation produced evaluations values, precision recall values. comparison result highest performance value 99.69%, 99.00%, 96.20%. In showed lowest 97.91%, 93.20%, 86.40%.