The Simalungun Batak script contained in the Batak Letter consists of several variants of forms depending on the language and region, in general there are five variants of the Batak Letter in Sumatra, namely Karo, Toba, Pakpak, Simalungun, Angkola Mandailing. This script is not widely known by the datu, namely people who are respected by the Batak community for mastering magic, fortune-telling, and calendaring so that errors still often occur in the introduction of this script, even though this script can be found in various libraries, namely the traditional books of the Batak community. By utilizing Artificial Neural Networks and using the Kohonen Self Organizing Map method which can be implemented in the Simalungun Batak Letter Recognition Application so that it can make it easier for the public to recognize the Simalungun Batak script. The results of the study that applied an artificial neural network with the kohonen self-organizing map algorithm made the recognition of the Simalungun Batak letter pattern easy for the public to recognize and could also preserve one of the cultures, namely the Simalungun Batak script and make it easier for users to implement it into the android-based application.

Javanese script is one of Indonesia's cultural heritages that are increasingly rarely used today. The difficulty of recognizing the shapes of letters, let alone writing them, is the main obstacle in using the Hanacaraka script. This research offers an alternative to Hanacaraka script recognition using a combination of image feature extraction and machine learning, where we utilize a pre-trained SquzeeNet model and Multilayer Backpropagation algorithm. Of the 18 models built using ReLu, Sigmoid, and Tanh activation functions, we found that the Tanh activation function, using the combination of 50-50-100 neuron configuration and 25 epochs, was the most optimal function used to classify the training data with accuracy, precision, and recall values of 93.8%. Meanwhile, the Tanh activation function, using the 50-100-50 neuron configuration and 50 epochs, is the most optimal function to classify the testing data, with accuracy, precision, and recall values of 89.1%, 89.5%, and 89.5%. All built models show a training and testing performance ratio below 10%. From this result, we conclude that all models have good reliability in the training and testing classification process.