Gå til hovedinnhold
0
Jump to main content

Online Machine Learning for Graph Topology Identification from Multiple Time Series

Bakht Zaman of the Faculty of Engineering and Science at the University of Agder has submitted his thesis entitled “Online Machine Learning for Graph Topology Identification from Multiple Time Series“ and will defend the thesis for the PhD-degree Thursday 29 October 2020. (Photo: Private)

In this PhD Dissertation, we address the challenges of streaming data, time-varying models, and missing data by proposing different online algorithms under the framework of online machine learning.

Bakht Zaman

PhD Candidate

The disputation will be held digitally, because of the Corona covid-19-situation. Spectators may follow the disputation digitally – link is available below.

 

Bakht Zaman of the Faculty of Engineering and Science at the University of Agder has submitted his thesis entitled “Online Machine Learning for Graph Topology Identification from Multiple Time Series“ and will defend the thesis for the PhD-degree Thursday 29 October 2020. 

He has followed the PhD Programme at the Faculty of Engineering and Science with Specialization in Information- and Communication Technology (ICT) at the University of Agder. His PhD project is funded by the PETROMAKS (Norwegian only) Smart-Rig grant 244205, see also Petromaks 2 - Large-scale Programme for Petroleum Research..

Summary of the Thesis by Bakht Zaman:

Online Machine Learning for Graph Topology Identification from Multiple Time Series

Multi-dimensional time series data are observed in many real-world systems, where some of the time series are influenced by other time series.

Identifying these relations encoded in a graph structure or topology among the time series is of paramount interest in certain applications. The inferred topology can provide insights about the underlying system and can assist in inference tasks such as prediction and anomaly detection.

Challenges in time series data

A major challenge in topology identification is that the assumption of static topology does not hold always in practice since most of the practical systems are evolving with time.

Another challenge is that the data is not available at once - it is coming in a streaming fashion. Hence, batch methods are not applicable due to their computationally complexity.

Moreover, time series data can contain missing values due to faulty sensors, privacy and security reasons, or due to saving energy.

Challenges addressed and dealt with

In this PhD Dissertation, we address the challenges of streaming data, time-varying models, and missing data by proposing different online algorithms under the framework of online machine learning.

The proposed algorithms can deal with streaming data, are computational and memory efficient, and can track the (possibly) time-varying topologies.

To evaluate the performance of the proposed online algorithms, convergence guarantees are derived.

Furthermore, numerical results based on synthetic and real data, which corroborate with the theoretical analysis, are presented and discussed in this dissertation.

 

Disputation facts: 

The trial lecture and the public defence will take place online, via the Zoom conferencing app (link below)

Head of Department, Folke Haugland, Department of Information and Communication Technology, Faculty of Engineering and Science, will chair the disputation.

The trial lecture at 15:15 hours
Public defence at 17:00 hours

 

Given topic for trial lecture"Covariance Estimation: Theory and Applications"

Thesis Title: «Online Machine Learning for Graph Topology Identification from Multiple Time Series»

Search for the thesis in AURA - Agder University Research Archive, a digital archive of scientific papers, theses and dissertations from the academic staff and students at the University of Agder.

The thesis is available here:

 

The Candidate: Bakht Zaman: (1990, Mardan, Pakistan) B.S degree in Electronics Engineering from COMSATS Institute of Information and Technology, Abbottabad, Pakistan (2012) and M.S degree from Faculty of Electrical Engineering, Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Topi, Pakistan (2015). In January 2016, he started his studies towards the PhD degree at WISENET Center, Department of ICT, University of Agder, Norway. 

Opponents:

First opponent: Professor Antonio Garcia Marques, University King Juan Carlos, Spain

Second opponent: Professor Antonio Ortega, University of Southern California, USA

Professor Christian Walter Peter Omlin, Department of ICT, University of Agder, is appointed as the administrator for the assessment commitee.

Supervisors were Professor Baltasar Beferull-Lozano, UiA (main supervisor), Associate Professor Daniel Romero, UiA and Postdoctoral Research Fellow Luis Miguel Lopez Ramos, UiA (co-supervisors)

 

 

What to do as an audience member:

The disputation is open to the public, but to follow the trial lecture and the public defence, which is transmitted via the Zoom conferencing app, you have to click in as an audience member.

We ask audience members to join the virtual trial lecture at 15:05 at the earliest and the public defense at 16:50 at the earliest. After these times, you can leave and rejoin the meeting at any time. Further, we ask audience members to turn off their microphone and camera and keep them turned off throughout the event. You do this at the bottom left of the image when in Zoom. We recommend you use ‘Speaker view’. You select that at the top right corner of the video window when in Zoom.

Opponent ex auditorio:

The chair invites members of the public to pose questions ex auditorio in the introduction to the public defense, with deadlines. Questions can be submitted to the chair Folke Haugland at e-mail folke.haugland@uia.no