PhD scholarships at the IT University of Copenhagen, in relation to the Danish AI Pioneer Centre

PhD scholarships at the IT University of Copenhagen, in relation to the Danish AI Pioneer Centre

The IT University of Copenhagen invites applications for one or more PhD scholarships in the context of the national Pioneer Centre for Artificial Intelligence (P1), specifically in the collaboratories called Extended Reality (XR), Networks and Graphs (NG), and Speech and Language (SL). Start date is flexible, in the early autumn of 2023.

The Pioneer Centre for AI is a multi-university national cooperation focusing on fundamental research within an interdisciplinary framework, contributing to solving society's greatest challenges. It is led by University of Copenhagen. The other members are Technical University of Denmark, IT University of Copenhagen, Aalborg University, and Aarhus University.

The IT University's vision is to create and share knowledge that is profound and leads to groundbreaking information technology and services for the benefit of humanity. The Computer Science Department seeks to conduct research and education that are both foundational and have impact on industry and society.

Job description

You are encouraged to align yourself with one of the concrete proposals and possible supervisors listed at the end of this call, but you can exceptionally instead propose your own PhD project within the goals of the AI Pioneer Centre and the above-mentioned collaboratories. We encourage early contact to potential supervisors.

General information

The IT University of Copenhagen (ITU) is a teaching and research-based tertiary institution concerned with information technology (IT) and the opportunities it offers. The IT University has more than 160 full-time Faculty members. Research and teaching in information technology span all academic activities which involve computers including computer science, information and media sciences, humanities and social sciences, business impact and the commercialization of IT.

Successful applicants will be employed and enrolled at the ITU University for a period of 3 or 4 years depending on the university degree level of the applicant.

Qualification Requirements

The following qualifications are required:

3-year programme: candidates must have an MSc degree (or equivalent).

4-year programme: candidates must have a BSc degree (or equivalent) plus one year of Master level studies.

Salary

Appointment and salary will be in accordance with the Ministry of Taxation's agreement with the Danish Confederation of Professional Associations (AC).

Application

Qualified applicants typically have an established research record within the potential supervisor's field. A competitive application must include:

A cover letter including the name of a suggested supervisor at ITU who is linked to the applicant's field of research.

An excellent project proposal (statement of purpose), which provides evidence of independent thinking, novelty and originality, and a state-of- the-art grasp of the targeted research field matching the interest and competences of a potential ITU supervisor (max. 5 pages).

Documentation of academic degree(s): diploma(s) and grade transcripts that documents outstanding academic achievement.

Curriculum Vitae. The CV should contain at least two academic/professional references.

List of Publications, if any. The list of publications should contain the applicant's relevant scientific publications in recognized, peer-reviewed venues, such as international conferences or journals in the suggested supervisor's field, or an excellent master's thesis.

Co-author statements for publications with more than one author. This can be in the form of a traditional co-author statement or as part of a recommendation letter.

Letters of recommendation, if any. Letters of recommendation should be of relevance to the assessment of your potential as a researcher.

Documentation of English language skills

Qualified applicants are expected to be on an excellent level of education. It is an advantage if the applicant has relevant academic experience as well, e.g. serving as Teaching Assistant, Research Assistant or writing publications.

It is understood that even highly qualified applicants will satisfy these requirements to varying degrees, for example because of differences in publication cultures in various research fields.

Applications without the above-mentioned required documents will not be assessed.

Qualified applicants are expected to have a high level of education. It is an advantage if the applicant has relevant academic experience as well, e.g., serving as Teaching Assistant or Research Assistant or writing publications.

It is understood that even highly qualified applicants will satisfy these requirements to varying degrees, for example because of differences in publication cultures in various research fields.

The applicant will be accessed according to the Appointment Order from the Ministry of Science, Technology and Innovation of 13 March 2012.

The IT University may use shortlisting in connection to the recruitment process. In case of shortlisting, the Chair of the hiring committee selects applicants for assessment in consultation with the hiring committee after the application deadline. All applicants are notified whether their application has been passed for assessment. The shortlisting of candidates for assessment is based on the criteria in the job posting.

Early contact with our faculty staff is encouraged.

Questions related to the Pioneer Centre context may be directed to department head for Computer Science, Peter Sestoft, [email protected].

Questions related to procedure may be directed to the PhD School ([email protected]) or to the HR department, Kristina Køj-Udsen, [email protected].

Application procedure

You can only apply for this position through our e-recruitment system. Apply by pressing the button "Apply for position" in the job announcement on our website.

Please read the guidelines for applicants carefully before filling in the application form. You can see the guidelines through this link.

Applications should be written in English. All other documents, such as transcripts should be accompanied by a translation in English or Danish.

Please note that all application material will be destroyed after the assessment.

Application deadline: 1 April 2023, at 23:59 CEST.

Applications/enclosures received at ITU after the application deadline will not be taken into consideration. If you submit an application, it is your responsibility to ensure that it arrives before the deadline so please allow sufficient time for upload of publications and other documents.

The IT University invites all qualified researchers regardless of age, gender, religious affiliation or ethnic background to apply for the positions.

PhD scholarships at the IT University of Copenhagen in relation to the Danish AI Pioneer Centre

The IT University of Copenhagen invites applications for one or more PhD scholarships in the context of the national Pioneer Centre for Artificial Intelligence (P1), specifically in the collaboratories called Extended Reality (XR), Networks and Graphs (NG), and Speech and Language (SL). Start date is flexible, in the early autumn of 2023.

The Pioneer Centre for AI is a multi-university national cooperation focusing on fundamental research within an interdisciplinary framework, contributing to solving society's greatest challenges. It is led by University of Copenhagen. The other members are Technical University of Denmark, IT University of Copenhagen, Aalborg University, and Aarhus University.

The IT University's vision is to create and share knowledge that is profound and leads to groundbreaking information technology and services for the benefit of humanity. The Computer Science Department seeks to conduct research and education that are both foundational and have impact on industry and society.

Job description

You are encouraged to align yourself with one of the concrete proposals and possible supervisors listed at the end of this call, but you can exceptionally instead propose your own PhD project within the goals of the AI Pioneer Centre and the above-mentioned collaboratories. We encourage early contact to potential supervisors.

General information

The IT University of Copenhagen (ITU) is a teaching and research-based tertiary institution concerned with information technology (IT) and the opportunities it offers. The IT University has more than 160 full-time Faculty members. Research and teaching in information technology span all academic activities which involve computers including computer science, information and media sciences, humanities and social sciences, business impact and the commercialization of IT.

Successful applicants will be employed and enrolled at the ITU University for a period of 3 or 4 years depending on the university degree level of the applicant.

Qualification Requirements

The following qualifications are required:

3-year programme: candidates must have an MSc degree (or equivalent).

4-year programme: candidates must have a BSc degree (or equivalent) plus one year of Master level studies.

Salary

Appointment and salary will be in accordance with the Ministry of Taxation's agreement with the Danish Confederation of Professional Associations (AC).

Application

Qualified applicants typically have an established research record within the potential supervisor's field. A competitive application must include:

A cover letter including the name of a suggested supervisor at ITU who is linked to the applicant's field of research.

An excellent project proposal (statement of purpose), which provides evidence of independent thinking, novelty and originality, and a state-of- the-art grasp of the targeted research field matching the interest and competences of a potential ITU supervisor (max. 5 pages).

Documentation of academic degree(s): diploma(s) and grade transcripts that documents outstanding academic achievement.

Curriculum Vitae. The CV should contain at least two academic/professional references.

List of Publications, if any. The list of publications should contain the applicant's relevant scientific publications in recognized, peer-reviewed venues, such as international conferences or journals in the suggested supervisor's field, or an excellent master's thesis.

Co-author statements for publications with more than one author. This can be in the form of a traditional co-author statement or as part of a recommendation letter.

Letters of recommendation, if any. Letters of recommendation should be of relevance to the assessment of your potential as a researcher.

Documentation of English language skills

Qualified applicants are expected to be on an excellent level of education. It is an advantage if the applicant has relevant academic experience as well, e.g. serving as Teaching Assistant, Research Assistant or writing publications.

It is understood that even highly qualified applicants will satisfy these requirements to varying degrees, for example because of differences in publication cultures in various research fields.

Applications without the above-mentioned required documents will not be assessed.

Qualified applicants are expected to have a high level of education. It is an advantage if the applicant has relevant academic experience as well, e.g., serving as Teaching Assistant or Research Assistant or writing publications.

It is understood that even highly qualified applicants will satisfy these requirements to varying degrees, for example because of differences in publication cultures in various research fields.

The applicant will be accessed according to the Appointment Order from the Ministry of Science, Technology and Innovation of 13 March 2012.

The IT University may use shortlisting in connection to the recruitment process. In case of shortlisting, the Chair of the hiring committee selects applicants for assessment in consultation with the hiring committee after the application deadline. All applicants are notified whether their application has been passed for assessment. The shortlisting of candidates for assessment is based on the criteria in the job posting.

Early contact with our faculty staff is encouraged.

Questions related to the Pioneer Centre context may be directed to department head for Computer Science, Peter Sestoft, [email protected].

Questions related to procedure may be directed to the PhD School ([email protected]) or to the HR department, Kristina Køj-Udsen, [email protected].

Application procedure

You can only apply for this position through our e-recruitment system. Apply by pressing the button "Apply for position" in the job announcement on our website.

Please read the guidelines for applicants carefully before filling in the application form. You can see the guidelines through this link.

Applications should be written in English. All other documents, such as transcripts should be accompanied by a translation in English or Danish.

Please note that all application material will be destroyed after the assessment.

Application deadline: 1 April 2023, at 23:59 CEST.

Applications/enclosures received at ITU after the application deadline will not be taken into consideration. If you submit an application, it is your responsibility to ensure that it arrives before the deadline so please allow sufficient time for upload of publications and other documents.

The IT University invites all qualified researchers regardless of age, gender, religious affiliation or ethnic background to apply for the positions.

PROJECT PROPOSALS

Project: The Business Value of Big Data Analytics and Machine Learning Algorithms

Speech & Language (SL) collaboratory

Recent years have seen massive increases in machine learning and deep learning techniques, such as artificial neural networks, for analysing large volumes of data. Most of the work so far has focused on optimizing the performance of specific algorithms and machine learning techniques and applying them to ever- larger quantities of data.

By applying large-scale simulation and data analysis of live business datasets, this PhD project investigates how much data and which machine learning algorithms are most effective and sustainable in creating value across different business decision-making contexts.

In doing so, the project will generate insights to inform decision-makers about the relevant scope of data and machine learning algorithms for specific business applications, thus making big data analytics more cost efficient, accessible, and sustainable. Research questions to be addressed include:

How big do data need to be to create value in business decision-making?

What is the relevant complexity of machine learning algorithms applied to business decision-making?

What is the most sustainable application of machine learning algorithms and big data analytics to different types of decision-making contexts?

By investigating business value of big data analytics, the project will contribute to existing data science research by moving beyond performance optimization on ever-larger quantities of data and suggesting more economic and appropriate applications of machine learning.

Expected PhD student qualifications:

Master's degree in Data Science, Information Systems, Innovation Management (with a focus on computational methods), or related fields.

Programming skills (focus on Python).

Ability to engage in interdisciplinary research combining data science, management, and information systems.

Supervisor: Jonas Valbjørn Andersen, [email protected], Business IT Department, IT University of Copenhagen.

PROJECT PROPOSALS

Project: Improving Efficiency of NLP Models by Exploiting NLP Datasets

Speech & Language (SL) collaboratory

Large language models have revolutionized the field of Natural Language Processing (NLP) since their introduction in 2018, leading to better performance on almost all NLP tasks. These models are first pre-trained on enormous amounts of raw, unlabelled texts, and then fine-tuned for the target task of interest. However, due to their computational costs, training these models can currently only be done by large tech companies. Recently, it has been shown that language models can be improved by an intermediate step of training on a diverse set of NLP tasks. This second step of training is computationally cheaper compared to the initial pre-training paradigm, with normally around 1,000 times smaller data training data (millions of words versus billions).

In this project we aim to investigate whether competitive performance can be obtained without the pre-training step. When training a language model only on a set of diverse NLP tasks, training costs will be only a fraction of current models. Sub-research questions that need to be tackled are:

How can we select NLP datasets that are diverse enough to lead to a versatile model?

How do we have to adapt the hyperparameters of the training procedure to the shorter training and more direct signal?

For which set of NLP tasks does the proposed approach lead to competitive performance?

Even if performance would be lower compared to current state-of-the-art models, the resulting model has the following benefits:

Less dependence on the few groups that have the capacity to train current models.

More control over training data: one problem with the enormous size of the training data for large language models is that it is impossible to curate.

Lower environmental impact: training BERT-base, a commonly used language model from 2018, has been shown to cost approximately 650 kg of CO2, while recent models are trained even much longer.

Better tuning: current large language models have many arbitrary design decisions.

Expected PhD student qualifications:

Master's degree in Natural Language Processing, Data Science, Linguistics (with a computational component) or related fields.

Programming skills (focus on Python)

Supervisor: Rob van der Goot, [email protected], Computer Science Department, IT University of Copenhagen; and possible co-supervisor.

Project: Modelling compositional language structures across languages

Speech & Language (SL) collaboratory

The origin of language development of human beings can at least be traced back to two sources: the survival needs for communication and the unique wiring of human brains. The two factors can co-develop with each other through lifespan. Nevertheless, languages from the whole world show tremendous diversity. Given the fact that all human brains share similar structures and connection patterns, which presumably afford similar ways to encode and decode information from the external world, there likely exist some hidden common structures, like some word orders or numbers of phonemes in a language, that are shared at least within, or even across, language families. An important constraint for these similar structures comes from the theory that the brain is processing language embodied in all our senses and via processing streams that are also involved in a range of other cognitive functions from specific motor control up to general problem-solving. This suggests that language comprehension and production, in fact, developed on top of existing information processing schemes, which in turn might have similarly shaped how the different language families have developed. A particular mechanism that was recently hypothesised to give rise to the structure of the brain's sequence processing is temporal compositionality and chunking, which seemingly operate on language sequences as well. With this PhD project, we want to identify and describe the specific, latent temporal encoding structures that may constrain the temporal features of spoken language.

In this project, the candidate will study structure patterns in spoken language and investigate how to build a model that can extract temporal characteristics of speech across different languages. Cornerstones of the project:

Large-scale audio data: The project will start with the analysis of existing audio data sets, in particular the audiobooks of 'Harry Potter' and 'The Little Prince' in multiple languages, including English, Danish, and Japanese and the aim to cover language families in broad range.

Spatio-temporal artificial neural networks (ANNs): A series of computational models from neuro-cognitively plausible cortex-level ANNs up to recent deep learning models of language processing will be developed and analysed. A starting point is recently suggested models that can explicitly learn the timescale parameters of temporal data.

Comparison with imaging data: To further verify if the extracted computation model reflects the processing dynamics in human brains, we plan to perform correlation analyses between the layers or clusters of neurons activating on certain timescales or on certain linguistic components found in the model and imaging data. Such data could stem from existing fMRI or EEG recordings or could be obtained in our lab within ITU.

Since the project is interdisciplinary, collaborations within the AI Pioneer Centre as well as with experts in computational neuroscience and developmental psychology in Germany and Japan are planned.

Expected PhD student qualifications:

Master's degree in Computer Science, Computational Neuroscience, Computational Linguistics, Developmental Psychology, Cognitive Science, or related fields.

Programming skills (such as in Python).

Strong interest in interdisciplinary research, at the intersection of Computer Science, Neuroscience, and Psychology, particularly between Natural Language Processing, Machine Learning, and Psycholinguistics.

Supervisor: Stefan Heinrich, [email protected], and Barbara Plank, [email protected], Computer Science Department, IT University of Copenhagen.

Project: Automatic Analysis of Mental Health by Machine Learning

Extended Reality (XR) collaboratory

Mental disorders often go undiagnosed or misdiagnosed for years, including bipolar disorder, depression, autism, etc., with negative impacts on individuals and communities.

This PhD project aims to facilitate better understanding of mental disorders, by investigating the audio and video data of patient vs. control group, using tailored machine learning methods and developing them further. The long-term goal is to potentially open the way to automatic diagnosis of mental disorders based on the new findings.

Previous studies have shown that facial expressions contain some information about the mental health of a person, specifically that the emotional response differs between people with bipolar disorder and the healthy control group, which offers great prospects for further investigation. Previous approaches have employed neural networks, e.g., LSTMs (Long-Short-Term-Memory), coupled with Multi-Layer Perceptron networks on videos to classify mental disorders. While the results are encouraging, they mostly focus on the classification or prediction task without investigating any further.

This project aims to go further by analysing different modalities of data, such as speech, eye movements, behaviour, virtual reality (VR) interactions. New Machine learning algorithms, tailored to one modality or combinations, should uncover new insights into the nature of mental disorders.

The project will include:

Extensive background investigation of state-of-the-art, specifically:

What kind of machine learning methods have been used to predict mental disorders?

Investigation which modalities are most relevant in those approaches.

Based on previous findings: can conclusions be made about properties of the disease or specific group differences, which have not been noted or further investigated so far?

Developing new Machine Learning methods to answer the following questions:

Which properties distinguish the people with mental disorders from healthy control?

Which properties are unique to people with a specific mental disorder?

What kind of different variants of specific mental disorders do exist, and how can this knowledge be used further?

Expected PhD student qualifications:

Master's degree in Computer Science, Data Science, Physics, Applied Mathematics, or related fields.

Programming skills (focus on Python)

Supervisors: Stella Grasshof, [email protected], and Sami Brandt, [email protected], Computer Science Department, IT University of Copenhagen.

Project: Uncovering the fundamental limits of ML and AI for predicting human behaviour

Networks & Graphs (NG) collaboratory

Is everything predictable given enough data and computational power? There have been plenty of optimistic claims by journalists, researchers, and companies about the possibility to predict different phenomena from epidemics, the stock market, crime and even wars. However, research has found these claims to be exaggerated, demonstrating that social systems are notoriously difficult to predict. Nonetheless claims about predictability are generally believed by policy makes and the public.

With the rise of algorithmic decision making and with automated systems mediating an increasingly larger part of our social, cultural, economic, and political interactions, it is vital to understand the limits of prediction and when predictive accuracies fall short of expectations. The overreaching goal of this proposal is to develop an empirical and theoretical understanding of predictability in social networks and human mobility. Are prediction limits determined by the size and bias present in datasets, the scale of computational power, or are there fundamental limits to prediction.

To tackle this question the project will focus on three subgoals. 1) We will build on top of existing statistical frameworks to predict events driven by collective phenomena in social systems. 2) Develop an understanding of how biases in data affect predictive accuracies. 3) Use information theoretical methodologies to quantify the upper limit of predictability. This project focuses on prediction in social systems and aims to develop a more critical understanding what predictive systems can be used for, and where we need to be careful.

Expected PhD qualifications:

Master's degree in Computer Science, Data Science, Physics, Applied Mathematics, or related fields, or in branches of Social Science relevant to the project (e.g., Psychology, Sociology)

Comprehensive programming skills (focus on Python), with a focus on data analysis

Motivation to do interdisciplinary research, at the intersection of data science and computational social science

This prospective PhD student will interact with the Networks, Data, and Society (NERDS) team at ITU, and with the Copenhagen Center for Social Data Science (SODAS) at KU. Both groups conduct quantitative research at the intersection of Data Science, Machine Learning, Computational Social Science, and Network Science. The two groups have collaborations with scholars from other major universities in Denmark (DTU, Aarhus, Aalborg), and internationally with universities and research centres in Europe and in the US, where the prospective PhD student can spend a research semester. Recently, the NERDS group won the awards for best research environment in Denmark.

Supervisor: Vedran Sekara, [email protected], Computer Science Department, IT University of Copenhagen (ITU), and co-supervisor Roberta Sinatra, Computer Science Department, IT University of Copenhagen and SODAS, University of Copenhagen (KU), [email protected].

IT University of Copenhagen, Rued Langgaards Vej 7, DK-2300 Copenhagen S

www. itu.dk