The area of ??machine learning and related deep learning has been revived in the past ten years. This was made possible by the increasing computational capabilities of computers and big data. In addition to academia there is in increasing trend in the use deep learning in companies such as IBM, Google, Facebook, etc. Among others, deep neural networks are used to identify objects in images, speech recognition, language modeling, companies use it in advertising, to study user behaviour and providing the best user experience.
Time and place: at 16:00 in FAMNIT-1-MP2
Lecturer: Lamb Čočorovski
Title: TensorFlow - a tool for modeling neural networks
With the proliferation of community and a growing interest in deep learning, a lot of modeling tools have been implemented and are available for users to use. Recently, Google's machine learning TensorFlow became open source. They use it in their products such as Google Photos, Google Translate, Google Search, speech recognition, etc. At the seminar we will learn more about TensorFlow, its structure and function. Moreover, we will describe convolutional neural networks and how to use them for the classification of images in TensorFlow for an Andrid app. In addition to this practical example the links between TensorFlow and the Android operating system will be presented.