The Rainfall Forecast Using Multiple Linear Regression and K-Nearest Neighbors with Parameters of Temperature, Air Humidity, Surface Pressure, Horizontal Wind Speed and Vertical Wind Speed

Abstract

ABSTRACT
Weather is a natural event that often occurs and closely coexists with human life. Weather in an area can be forecasted by calculating several factors that cause weather changes, including air temperature, air humidity, surface pressure, horizontal wind speed, and vertical wind speed, these factors are a series that can affect rainfall in an area. The prediction in this study uses the Multiple Linear Regression method and the K-Nearest Neighbor in predicting the occurrence of rainfall in the Brebes Regency area. Often the methods used in predicting rainfall provide data that is still not maximal in predicting the possibility of rainfall occurring. The results of this study indicate that the use of Multiple Linear Regression and K-Nearest Neighbor has a Mean Absolute Error level of 0.005 and 0.16 which concludes that K-Nearest Neighbor is more capable of making more accurate predictions of the rainfall that will occur than Multiple Linear Regression.
Keywords: Weather, Temperature, Humidity, Multiple Linear Regression, K-Nearest Neighbor