Mezun Etkinlikleri 2 - Yeni Nesil, Makine Öğrenimi Tabanlı Mobil Ağ Otomasyonunun Temelleri - Yakup Tarık KRANDA

Bölümümüz mezunlarının tecrübelerinin yeni nesile aktarılması ve potansiyel iş dallarının ve bilişim dünyasının eğilimlerini görmek adına öğrencilerimize yönelik olarak mezunlarımızın katılımı ile başladığımız etkinlik serisinin ikincisini 5 Nisan 2021 Pazartesi 11:00 da gerçekleştiriyoruz. 

Bu etkinliğimiz 2005 yılı mezunumuz P.I. Works'de Product and Applied Research Group Manager olarak çalışan Yakup Tarık KRANDA tarafından "Yeni Nesil, Makine Öğrenimi Tabanlı Mobil Ağ Otomasyonunun Temelleri" başlığında verilecektir. Konu kapsamı dışında da öğrencilerimizin farklı alanlardaki soruları da cevaplandırılacaktır.

Etkinliğimizin konu başlığı ile ilgili detay bilgiler ve 5 Nisan 11:00'da başlayacak çevrimiçi toplantı için bağlantı bilgisi aşağıdadır. 

Etkinliğe katılmak için bu bağlantıyı ziyaret ediniz. Etkinliğimiz Microsoft Teams ortamı üzerinden yapılmaktadır.

Yakup Tarik Kranda is a Lead Product Manager and Researcher at P.I. Works HQ office, responsible for productization of next generation network automation solutions leveraged by artificial intelligence and advanced optimization technologies.  He received his BSc. and MSc. valedictorian degrees in Computer Engineering from Istanbul University and is a dissertation stage PhD. candidate at the same department focusing on operational efficiency of wireless network optimization with AI/ML. He also holds a second BSc. degree in Industrial Engineering from Istanbul University. He is author of 10 patents in Telco domain and has been contributing to ITU-T Machine Learning Focus Group workshops and technical reports. He started his professional career during his undergraduate years as co-founder of Digiport Software and part-time software-engineer at Microsoft Turkey between 2005 and 2007. After working as Research Assistant at Istanbul University between 2005 and 2008, he continued his career at Innova A.Ş as senior engineer and software architect focusing on Türk Telekom VAS services design until 2010. He then worked for Turkcell Technology R&D Group as Software Architect, Product Manager and Corporate Entrepreneurship Supervisor, before joining P.I. Works in 2015.

Fundamentals of Next Generation, Machine Learning Based Mobile Network Automation
Mobile networks have been experiencing significant data demand for more than a decade due to increased data connectivity in every dimension of life with mobile applications and services. To meet the expected Quality of Experience (QoE) per service, new technologies are revealed including 5G new radio, slicing, network virtualization, Software-Defined Networking (SDN) and so on. All these advancements in technology that bring more capacity and flexibility to manage the scarce resources, require more end-to-end advanced automation solutions i.e. Self-Organizing Network (SON) to handle increased complexity. Key requirement to achieve automation is to feed more data from multiple domains, providing prediction and correlation capabilities, being able to define normal and abnormal, apply preventive actions proactively by using historical patterns and spatial correlations. All these key requirements will shape the architecture of mobile networks to optimize the cost and network capacity.

In this lecture we will cover the topics below.

  • What are the main components of contemporary Mobile Network (MN) architecture?

  • How should a next generation MN automation be organized to meet all expectations and embrace these advancements in technology?

  • What types of machine learning applications do exist in Telco domain?

  • What is the general flow of machine learning projects that we track in applied research department?

  • What topics that you need to be careful when considering an ML based solution?

  • What are the fundamental machine learning algorithms that are used in industry and academia for mobile network automation?

  • Some use-cases and their results from ML based network automation.


Yaklaşan Etkinlikler

https://igccc.info/

Konum