Penerapan Learning Vector Quantization (LVQ) Untuk Klasifikasi Data Citra Digital

The rapid development of information technology has increased the utilization of digital images in various fields, creating a need for classification methods that are accurate and efficient. One method that can be applied to classify numerical data obtained from image feature extraction is Learning Vector Quantization (LVQ). This study aims to implement the LVQ method for digital image classification based on numerical features and to evaluate its performance in terms of accuracy. The data used in this study consist of grayscale digital images that have undergone a feature extraction process and are represented as numerical vectors. The dataset is divided into two classes, namely Class A and Class B. The research stages include data collection, grayscale conversion, feature extraction, LVQ training, and classification testing. The classification results are evaluated using a confusion matrix and accuracy measurement. The experimental results show that the LVQ method successfully classified all test data correctly, achieving an accuracy rate of 100%. These results indicate that Learning Vector Quantization is an effective method with good performance for classifying digital image data based on numerical features.