Prediction Of Used Car Prices Using K-Nearest Neighbour, Random Forest And Adaptive Boosting Algorithm

Abstract

In the midst of busy  society and high lifestyle,  there are now many car offerings with advanced features.  The more sophisticated a car is, the  price  increases. This  makes people prefer  to buy a used car with specifications that are still suitable for use.  Therefore  , used car entrepreneurs try to provide prices that are in accordance with the  quality of the  car. In order for  the price  of the   used car offered to  be in accordance with the market  and not make used car entrepreneurs    suffer losses,  it  is necessary to predict  the right and accurate price.    This study aims to help used car  business owners to determine the appropriate  car price  using 3 algorithms, namely K-nearest neighbor, Random Forest and AdaBoost. The novelty of this study is the improvement in the accuracy of the prediction model of a single model.  The results of  this study are that the algorithm that has the best performance is Random Forest. This is shown by the smallest MSE and RMSE values among others. The MSE value is 117.142273 and the RMSE value is below 1 which is 0.342261.