Currently Faculty of Technologies is hosting dr. Georgios Ouzounis. Georgios is a computer vision/machine learning engineer specializing in advanced, high performance image/video/data analytics, processing and feature extraction algorithms, cloud processing and distributed systems. Fields of his expertise include artificial, convolutional and recurrent neural networks for 2D/3D deep learning, self-organizing maps and auto-encoders. He has doctorate degree in mathematics and natural sciences. Georgios works as Vice President (Data Science) at ElectrifAi.

Georgios Ouzounis was invited to deliver the 5-day intensive course Crash Course in Deep Learning in the period from September 16 to 20, 2019. The aim of this course is to familiarize students with notions and concepts from the fields of data science and machine- & deep-learning, and to showcase how to build, train and deploy artificial neural networks and other advanced derivative architectures for solving a wide range of real-world challenges. The skill-sets developed upon completion of the course would allow attendants to engage directly in data-analytics problem solving and enrich their curriculum with a ‘must-have’ for the global job market.

The course attendees are the students who study Infotronics, Information Finance Systems, Automatic Control, Software Systems and Administration of Computer Networks.

Deep-Learning is a thriving, state-of-the-art field of data analytics that shapes developments in almost every field of the modern global economy. It is the driving technology in data science and artificial intelligence and has a very broad field of applications. Indicatively, the Deep-Learning market was worth USD 2.28 Billion in 2017 and is expected to reach USD 18.16 Billion by 2023, at a CAGR of 41.7 % from 2018 to 2023.

 

Schedule of lectures

Date Time Content Location
September 16 12.40–16.10 1. Course Introduction
2. Computational Infrastructures & Resources
3. Tutorial Session 1 – Github I/O
4. Tutorial Session 2 – NumPy, Pandas, MatPlotLib
1-61 (Pramonės pr. 20)
September 17 9.00–12.30 1. Features
2. Tutorial on Feature Selection
3. Cross Validation
4. Model Evaluation
1-61 (Pramonės pr. 20)
September 18 11.45–15.15 1. Artificial Neural Networks – Theory part 1
2. Artificial Neural Networks – Theory part 2
3. Artificial Neural Networks – Build, Test and Deploy
4. Artificial Neural Networks – Model Evaluation, Tuning and Improvement
1-61 (Pramonės pr. 20)
September 19 9.00–12.30 1. Convolutional Neural Networks – Theory Part 1
2. Convolutional Neural Networks – Theory part 2
3. Convolutional Neural Networks – Build, Test and Deploy
4. Convolutional Neural Networks – Practical Exercise – Improve performance
1-61 (Pramonės pr. 20)
September 20 9.00–12.30 1. Recurrent Neural Networks – Theory part 1
2. Recurrent Neural Networks – Theory part 2
3. Recurrent Neural Networks – Build, Test and Deploy
4. Recurrent Neural Networks – Model Evaluation, Tuning and Improvement
1-61 (Pramonės pr. 20)