Perbandingan Algoritma Xception dan VGG16 Untuk Pengenalan Lebah Pollen-Bearing
Having scheduled observations will help beekeepers know about bee diseases, bee hive health and poisons that may be carried by bees. If this can be done with the help of a computer, it will reduce the time and cost of beekeeping. In addition, honey and nest production will increase both in terms of quality and quantity. This study aims to analyze the performance of the Xception and VGG16 algorithms for the recognition of pollen-bearing bees. In the experimental results above, the VGG16 model with fine_tuning obtained the best testing accuracy value of 83.33%. Likewise, with the best Cohens kappa, F1_score, ROC AUC, Precision, and Recall values obtained by the VGG16 model with fine_tuning. For the Xception model, the best obtained without fine tuning is 72.22%. From the experimental results, it is concluded that the pre-trained VGG16 model with fine_tuning is more suitable for use in the bee_pollen dataset compared to the Xception model, either with fine_tuning or without fine_tuning.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.