CD Skripsi
Prediksi Jumlah Kecelakaan Dan Kerugian Materiil Di Provinsi Riau Menggunakan Metode Support Vector Regression
Traffic accidents are things that often cause loss of life. The number of traffic accidents varies every year, therefore anticipation is needed such as the addition of personnel on each road that is prone to accidents. This study aims to predict the number of accidents and material losses in Riau Province using the Support vector regression (SVR) method. The traffic accident data obtained is analyzed and modeled with SVR optimized using Algorithms to improve prediction accuracy. The results showed that the optimized SVR model had an average Root Means Squared Error (RMSE) error rate for the predicted number of accidents was 3.22, while the material loss RMSE result was 37,459,129. The process of epsilon and cost, has been shown to be effective in improving model performance. The selection of the optimal value for each parameter is carried out using the fold cross validation technique. Visualization of prediction results is presented through a web-based dashboard using RShiny, allowing interactive analysis of accident data in each Police Station in Riau Province. The results of this study can be used as a basis in planning strategies to reduce material losses due to traffic accidents.
Keywords : Material loss, Prediction Of The Number Of Accidents, RMSE, SVR, Tuning Parameters
Tidak tersedia versi lain