PERBANDINGAN COSINE SIMILARITY DAN K-NEAREST NEIGHBOUR PADA CASE BASED REASONING DALAM MENDETEKSI MASALAH SCALE PADA PRODUKSI MINYAK BUMI
Daftar Isi:
- Case Based Reasoning is an expert system method that can be used for detection by measuring the level of similarity between new cases and existing cases. In this study, researchers are interested in conducting a comparative analysis of the results of scale detection on petroleum production using Cosine Similarity and K-Nearest Neighbor. On the results of the detection of the expert system using Cosine Similarity, the accuracy value is 87%. Different results were obtained from the expert system using K-Nearest Neighbor with different k values, 100% accuracy value at k = 3, 87% at k = 5 and k = 7, 80% at k = 11, k = 15 and k = 17, and 67% at k = 23.