20-21 September 2017, BULGARIA 1 Proceedings of the International Conference on Information Technologies (InfoTech-2017) 20-21 September 2017, Bulgaria MOBILE APPLICATION FOR PATTERN RECOGNITION BASED ON MACHINE LEARNING Desislava Ivanova, Vladimir Kadurin Technical University of Sofia, Faculty of Applied Mathematics and Informatics e-mails: d_ivanova@tu-sofia.bg, vladimirkadurin@gmail.com Bulgaria Abstract: The paper presents the development of mobile application based on artificial intelligence. The functionality of the proposed mobile application is to recognize handwritten objects (letters and digits). For the development of the mobile application, the Artificial Neural Network (ANN) is used. The application is implemented on iOS platform, written in programming language Swift. Key words: Mobile Application, Pattern Recognition, Artificial Neural Network (ANN), iOS. 1. INTRODUCTION Artificial intelligence is very popular nowadays. Terms like artificial intelligence are very common. Artificial intelligence is not a new science. It appears in Greek mythology, showing the human desire for human alike machines, which can react autonomously to outside stimuli. But it is fair to be said that artificial intelligence begins with the invention of the first computer in 20th century [1, 2]. Artificial intelligence is very broad term, which includes many areas. One of them is machine learning. It is a science for algorithms, which can learn from different data sets and to make conclusion based on them. Machine learning has three main categories:  Supervised Learning - algorithms learn from samples, which we prepare in advance. Then they use these samples as analogy for evaluation of new unknown data.  Unsupervised Learning - algorithms learn from random data, which is not prepared in advance, and discover their own structure and meaning of the data.