# Available Master's thesis topics in machine learning

## Main content

Here we list topics that are available. You may also be interested in our list of completed Master's theses.

## Approximation algorithms for learning Bayesian networks

Bayesian networks are probabilistic models that are used to represent multivariate distributions. The core of a Bayesian network is its structure, a directed acyclic graph (DAG), that expresses conditional independencies between variables.

Typically, the structure is learned from data. The problem is NP-hard and thus exact algorithms do not scale up and one often resorts to heuristics that do not give any quality guarantees.

A recent paper presented a moderately exponential time approximation algorithm that can be used to trade between running time and quality of the approximation. However, the paper is fully theoretical and we do not know whether the proposed algorithm is useful in practice.

Task: Implement the algorithm and do experiments to assess its practical performance.

Advisor: Pekka Parviainen

## Learning and inference with large Bayesian networks

Most learning and inference tasks with Bayesian networks are NP-hard. Therefore, one often resorts to using different heuristics that do not give any quality guarantees.

Task: Evaluate quality of large-scale learning or inference algorithms empirically.

Advisor: Pekka Parviainen

## Sum-product networks

Traditionally, probabilistic graphical models use a graph structure to represent dependencies and independencies between random variables. Sum-product networks are a relatively new type of a graphical model where the graphical structure models computations and not the relationships between variables. The benefit of this representation is that inference (computing conditional probabilities) can be done in linear time with respect to the size of the network.

Potential thesis topics in this area: a) Compare inference speed with sum-product networks and Bayesian networks. Characterize situations when one model is better than the other. b) Learning the sum-product networks is done using heuristic algorithms. What is the effect of approximation in practice?

Advisor: Pekka Parviainen

## Bayesian Bayesian networks

The naming of Bayesian networks is somewhat misleading because there is nothing Bayesian in them per se; A Bayesian network is just a representation of a joint probability distribution. One can, of course, use a Bayesian network while doing Bayesian inference. One can also learn Bayesian networks in a Bayesian way. That is, instead of finding an optimal network one computes the posterior distribution over networks.

Task: Develop algorithms for Bayesian learning of Bayesian networks (e.g., MCMC, variational inference, EM)

Advisor: Pekka Parviainen

## Large-scale (probabilistic) matrix factorization

The idea behind matrix factorization is to represent a large data matrix as a product of two or more smaller matrices.They are often used in, for example, dimensionality reduction and recommendation systems. Probabilistic matrix factorization methods can be used to quantify uncertainty in recommendations. However, large-scale (probabilistic) matrix factorization is computationally challenging.

Potential thesis topics in this area: a) Develop scalable methods for large-scale matrix factorization (non-probabilistic or probabilistic), b) Develop probabilistic methods for implicit feedback (e.g., recommmendation engine when there are no rankings but only knowledge whether a customer has bought an item)

Advisor: Pekka Parviainen

## Bayesian deep learning

Standard deep neural networks do not quantify uncertainty in predictions. On the other hand, Bayesian methods provide a principled way to handle uncertainty. Combining these approaches leads to Bayesian neural networks. The challenge is that Bayesian neural networks can be cumbersome to use and difficult to learn.

The task is to analyze Bayesian neural networks and different inference algorithms in some simple setting.

Advisor: Pekka Parviainen

## Deep learning for combinatorial problems

Deep learning is usually applied in regression or classification problems. However, there has been some recent work on using deep learning to develop heuristics for combinatorial optimization problems; see, e.g., [1] and [2].

Task: Choose a combinatorial problem (or several related problems) and develop deep learning methods to solve them.

References: [1] Vinyals, Fortunato and Jaitly: Pointer networks. NIPS 2015. [2] Dai, Khalil, Zhang, Dilkina and Song: Learning Combinatorial Optimization Algorithms over Graphs. NIPS 2017.

Advisors: Pekka Parviainen, Ahmad Hemmati

## Estimating the number of modes of an unknown function

Mode seeking considers estimating the number of local maxima of a function f. Sometimes one can find modes by, e.g., looking for points where the derivative of the function is zero. However, often the function is unknown and we have only access to some (possibly noisy) values of the function.

In topological data analysis, we can analyze topological structures using persistent homologies. For 1-dimensional signals, this can translate into looking at the birth/death persistence diagram, i.e. the birth and death of connected topological components as we expand the space around each point where we have observed our function. These observations turn out to be closely related to the modes (local maxima) of the function. A recent paper [1] proposed an efficient method for mode seeking.

In this project, the task is to extend the ideas from [1] to get a probabilistic estimate on the number of modes. To this end, one has to use probabilistic methods such as Gaussian processes.

[1] U. Bauer, A. Munk, H. Sieling, and M. Wardetzky. Persistence barcodes versus Kolmogorov signatures: Detecting modes of one-dimensional signals. Foundations of computational mathematics17:1 - 33, 2017.

Advisors: Pekka Parviainen, Nello Blaser

## Causal Abstraction Learning

We naturally make sense of the world around us by working out causal relationships between objects and by representing in our minds these objects with different degrees of approximation and detail. Both processes are essential to our understanding of reality, and likely to be fundamental for developing artificial intelligence. The first process may be expressed using the formalism of structural causal models, while the second can be grounded in the theory of causal abstraction [1].

This project will consider the problem of learning an abstraction between two given structural causal models. The primary goal will be the development of efficient algorithms able to learn a meaningful abstraction between the given causal models.

[1] Rubenstein, Paul K., et al. "Causal consistency of structural equation models." arXiv preprint arXiv:1707.00819 (2017).

Advisor: Fabio Massimo Zennaro

## Causal Bandits

"Multi-armed bandit" is an informal name for slot machines, and the formal name of a large class of problems where an agent has to choose an action among a range of possibilities without knowing the ensuing rewards. Multi-armed bandit problems are one of the most essential reinforcement learning problems where an agent is directly faced with an exploitation-exploration trade-off.

This project will consider a class of multi-armed bandits where an agent, upon taking an action, interacts with a causal system [1]. The primary goal will be the development of learning strategies that takes advantage of the underlying causal system in order to learn optimal policies in a shortest amount of time.

[1] Lattimore, Finnian, Tor Lattimore, and Mark D. Reid. "Causal bandits: Learning good interventions via causal inference." Advances in neural information processing systems 29 (2016).

Advisor: Fabio Massimo Zennaro

## Causal Modelling for Battery Manufacturing

Lithium-ion batteries are poised to be one of the most important sources of energy in the near future. Yet, the process of manufacturing these batteries is very hard to model and control. Optimizing the different phases of production to maximize the lifetime of the batteries is a non-trivial challenge since physical models are limited in scope and collecting experimental data is extremely expensive and time-consuming [1].

This project will consider the problem of aggregating and analyzing data regarding a few stages in the process of battery manufacturing. The primary goal will be the development of algorithms for transporting and integrating data collected in different contexts, as well as the use of explainable algorithms to interpret them.

[1] Niri, Mona Faraji, et al. "Quantifying key factors for optimised manufacturing of Li-ion battery anode and cathode via artificial intelligence." Energy and AI 7 (2022): 100129.

Advisor: Fabio Massimo Zennaro, Mona Faraji Niri

## Reinforcement Learning for Computer Security

The field of computer security presents a wide variety of challenging problems for artificial intelligence and autonomous agents. Guaranteeing the security of a system against attacks and penetrations by malicious hackers has always been a central concern of this field, and machine learning could now offer a substantial contribution. Security capture-the-flag simulations are particularly well-suited as a testbed for the application and development of reinforcement learning algorithms [1].

This project will consider the use of reinforcement learning for the preventive purpose of testing systems and discovering vulnerabilities before they can be exploited. The primary goal will be the modelling of capture-the-flag challenges of interest and the development of reinforcement learning algorithms that can solve them.

[1] Erdodi, Laszlo, and Fabio Massimo Zennaro. "The Agent Web Model--Modelling web hacking for reinforcement learning." arXiv preprint arXiv:2009.11274 (2020).

Advisor: Fabio Massimo Zennaro, Laszlo Tibor Erdodi

## Approaches to AI Safety

The world and the Internet are more and more populated by artificial autonomous agents carrying out tasks on our behalf. Many of these agents are provided with an objective and they learn their behaviour trying to achieve their objective as better as they can. However, this approach can not guarantee that an agent, while learning its behaviour, will not undertake actions that may have unforeseen and undesirable effects. Research in AI safety tries to design autonomous agent that will behave in a predictable and safe way [1].

This project will consider specific problems and novel solution in the domain of AI safety and reinforcement learning. The primary goal will be the development of innovative algorithms and their implementation withing established frameworks.

[1] Amodei, Dario, et al. "Concrete problems in AI safety." arXiv preprint arXiv:1606.06565 (2016).

Advisor: Fabio Massimo Zennaro

## Reinforcement Learning for Super-modelling

Super-modelling [1] is a technique designed for combining together complex dynamical models: pre-trained models are aggregated with messages and information being exchanged in order synchronize the behavior of the different modles and produce more accurate and reliable predictions. Super-models are used, for instance, in weather or climate science, where pre-existing models are ensembled together and their states dynamically aggregated to generate more realistic simulations.

This project will consider how reinforcement learning algorithms may be used to solve the coordination problem among the individual models forming a super-model. The primary goal will be the formulation of the super-modelling problem within the reinforcement learning framework and the study of custom RL algorithms to improve the overall performance of super-models.

[1] Schevenhoven, Francine, et al. "Supermodeling: improving predictions with an ensemble of interacting models." Bulletin of the American Meteorological Society 104.9 (2023): E1670-E1686.

Advisor: Fabio Massimo Zennaro, Francine Janneke Schevenhoven

## Multilevel Causal Discovery

Modelling causal relationships between variables of interest is a crucial step in understanding and controlling a system. A common approach is to represent such relations using graphs with directed arrows discriminating causes from effects.

While causal graphs are often built relying on expert knowledge, a more interesting challenge is to learn them from data. In particular, we want to consider the case where data might have been collected at multiple levels, for instance, with sensor with different resolutions. In this project we want to explore how these heterogeneous data can help the process of inferring causal structures.

[1] Anand, Tara V., et al. "Effect identification in cluster causal diagrams." Proceedings of the 37th AAAI Conference on Artificial Intelligence. Vol. 82. 2023.

Advisor: Fabio Massimo Zennaro, Pekka Parviainen

## Manifolds of Causal Models

Modelling causal relationships is fundamental in order to understand real-world systems. A common formalism is offered by structural causal models (SCMs) which represent these relationships graphical. However, SCMs are complex mathematical objects entailing collections of different probability distributions.

In this project we want to explore a differential geometric perspective on structural causal models [1]. We will model an SCM and the probability distributions it generates in terms of manifold, and we will study how this modelling encodes causal properties of interest and how relevant quantities may be computed in this framework.

[1] Dominguez-Olmedo, Ricardo, et al. "On data manifolds entailed by structural causal models." International Conference on Machine Learning. PMLR, 2023.

Advisor: Fabio Massimo Zennaro, Nello Blaser

## Topological Data Analysis on Simulations

Complex systems and dynamics may be hard to formalize in a closed form, and they can often be better studied through simulations. Social systems, for instance, may be reproduced by instantiating simple agents whose interactions generate complex and emergent dynamics. Still, analyzing the behaviours arising from these interactions is not trivial.

In this project we will consider the use of topological data analysis for categorizing and understanding the behaviour of agents in agent-based models [1]. We will analyze the insights and the limitations of exisiting algorithms, as well as consider what dynamical information may be glimpsed through such an analysis.

[1] Swarup, Samarth, and Reza Rezazadegan. "Constructing an Agent Taxonomy from a Simulation Through Topological Data Analysis." Multi-Agent-Based Simulation XX: 20th International Workshop, MABS 2019, Montreal, QC, Canada, May 13, 2019, Revised Selected Papers 20. Springer International Publishing, 2020.

Advisor: Fabio Massimo Zennaro, Nello Blaser

## Abstraction for Epistemic Logic

Weighted Kripke models constitute a powerful formalism to express the evolving knowledge of an agent; it allows to express known facts and beliefs, and to recursively model the knowledge of an agent about another agent. Moreover, such relations of knowledge can be given a graphical expression using suitable diagrams on which to perform reasoning. Unfortunately, such graphs can quickly become very large and inefficient to process.

This project consider the reduction of epistemic logic graph using ideas from causal abstraction [1]. The primary goal will be the development of ML models that can learn to output small epistemic logic graph still satisfying logical and consistency constraints.

[1] Zennaro, Fabio Massimo, et al. "Jointly learning consistent causal abstractions over multiple interventional distributions." Conference on Causal Learning and Reasoning. PMLR, 2023

Advisor: Fabio Massimo Zennaro, Rustam Galimullin

## Optimal Transport for Public Transportation

Modelling public transportation across cities is critical in order to improve viability, provide reliable services and increase reliance on greener form of mass transport. Yet cities and transportation networks are complex systems and modelling often has to rely on incomplete and uncertain data.

This project will start from considering a concrete challenge in modelling commuter flows across the city of Bergen. In particular, it will consider the application of the mathematical framework of optimal transport [1] to recover statistical patterns in the usage of the main transportation lines across different periods.

[1] Peyré, Gabriel, and Marco Cuturi. "Computational optimal transport: With applications to data science." Foundations and Trends in Machine Learning 11.5-6 (2019): 355-607.

Advisor: Fabio Massimo Zennaro

## Finalistic Models

The behavior of an agent may be explained both in causal terms (what has caused a certain behavior) or in finalistic terms (what aim justifies a certain behaviour). While causal reasoning is well explained by different mathematical formalism (e.g., structural causal models), finalistic reasoning is still object of research.

In this project we want to explore how a recently-proposed framework for finalistic reasoning [1] may be used to model intentions and counterfactuals in a causal bandit setting, or how it could be used to enhance inverse reinforcement learning.

[1] Compagno, Dario. "Final models: A finalistic interpretation of statistical correlation." arXiv preprint arXiv:2310.02272 (2023).

Advisor: Fabio Massimo Zennaro, Dario Compagno

## Automatic hyperparameter selection for isomap

Isomap is a non-linear dimensionality reduction method with two free hyperparameters (number of nearest neighbors and neighborhood radius). Different hyperparameters result in dramatically different embeddings. Previous methods for selecting hyperparameters focused on choosing one optimal hyperparameter. In this project, you will explore the use of persistent homology to find parameter ranges that result in stable embeddings. The project has theoretic and computational aspects.

Advisor: Nello Blaser

## Topological Ancombs quartet

This topic is based on the classical Ancombs quartet and families of point sets with identical 1D persistence (https://arxiv.org/abs/2202.00577). The goal is to generate more interesting datasets using the simulated annealing methods presented in (http://library.usc.edu.ph/ACM/CHI%202017/1proc/p1290.pdf). This project is mostly computational.

Advisor: Nello Blaser

## Persistent homology vectorization with cycle location

There are many methods of vectorizing persistence diagrams, such as persistence landscapes, persistence images, PersLay and statistical summaries. Recently we have designed algorithms to in some cases efficiently detect the location of persistence cycles. In this project, you will vectorize not just the persistence diagram, but additional information such as the location of these cycles. This project is mostly computational with some theoretic aspects.

Advisor: Nello Blaser

## Divisive covers

Divisive covers are a divisive technique for generating filtered simplicial complexes. They original used a naive way of dividing data into a cover. In this project, you will explore different methods of dividing space, based on principle component analysis, support vector machines and k-means clustering. In addition, you will explore methods of using divisive covers for classification. This project will be mostly computational.

Advisor: Nello Blaser

## Learning Acquisition Functions for Cost-aware Bayesian Optimization

This is a follow-up project of an earlier Master thesis that developed a novel method for learning Acquisition Functions in Bayesian Optimization through the use of Reinforcement Learning. The goal of this project is to further generalize this method (more general input, learned cost-functions) and apply it to hyperparameter optimization for neural networks.

Advisors: Nello Blaser, Audun Ljone Henriksen

## Stable updates

This is a follow-up project of an earlier Master thesis that introduced and studied empirical stability in the context of tree-based models. The goal of this project is to develop stable update methods for deep learning models. You will design sevaral stable methods and empirically compare them (in terms of loss and stability) with a baseline and with one another.

Advisors: Morten Blørstad, Nello Blaser

## Multimodality in Bayesian neural network ensembles

One method to assess uncertainty in neural network predictions is to use dropout or noise generators at prediction time and run every prediction many times. This leads to a distribution of predictions. Informatively summarizing such probability distributions is a non-trivial task and the commonly used means and standard deviations result in the loss of crucial information, especially in the case of multimodal distributions with distinct likely outcomes. In this project, you will analyze such multimodal distributions with mixture models and develop ways to exploit such multimodality to improve training. This project can have theoretical, computational and applied aspects.

Advisor: Nello Blaser

## Wet area segmentation for rivers

NORCE LFI is working on digitizing wetted areas in rivers. You will apply different machine learning techniques for distinguishing water bodies (rivers) from land based on drone aerial (RGB) pictures. This is important for water management and assessing effects of hydropower on river ecosystems (residual flow, stranding of fish and spawning areas). We have a database of approximately 100 rivers (aerial pictures created from totally ca. 120.000 single pictures with Structure from Motion, single pictures available as well) and several of these rivers are flown at 2-4 different discharges, taken in different seasons and with different weather patterns. For ca. 50 % of the pictures the wetted area is digitized for training (GIS shapefile), most (>90 % of single pictures) cover water surface and land. Possible challenges include shading, reflectance from the water surface, different water/ground colours and wet surfaces on land. This is an applied topic, where you will try many different machine learning techniques to find the best solution for the mapping tasks by NORCE LFI.

Advisor: Nello Blaser, Sebastian Franz Stranzl

## Learning a hierarchical metric

Often, labels have defined relationships to each other, for instance in a hierarchical taxonomy. E.g. ImageNet labels are derived from the WordNet graph, and biological species are taxonomically related, and can have similarities depending on life stage, sex, or other properties.

ArcFace is an alternative loss function that aims for an embedding that is more generally useful than softmax. It is commonly used in metric learning/few shot learning cases.

Here, we will develop a metric learning method that learns from data with hierarchical labels. Using multiple ArcFace heads, we will simultaneously learn to place representations to optimize the leaf label as well as intermediate labels on the path from leaf to root of the label tree. Using taxonomically classified plankton image data, we will measure performance as a function of ArcFace parameters (sharpness/temperature and margins -- class-wise or level-wise), and compare the results to existing methods.

Advisor: Ketil Malde (ketil.malde@hi.no)

## Self-supervised object detection in video

One challenge with learning object detection is that in many scenes that stretch off into the distance, annotating small, far-off, or blurred objects is difficult. It is therefore desirable to learn from incompletely annotated scenes, and one-shot object detectors may suffer from incompletely annotated training data.

To address this, we will use a region-propsal algorithm (e.g. SelectiveSearch) to extract potential crops from each frame. Classification will be based on two approaches: a) training based on annotated fish vs random similarly-sized crops without annotations, and b) using a self-supervised method to build a representation for crops, and building a classifier for the extracted regions. The method will be evaluated against one-shot detectors and other training regimes.

If successful, the method will be applied to fish detection and tracking in videos from baited and unbaited underwater traps, and used to estimate abundance of various fish species.

See also: Benettino (2016): https://link.springer.com/chapter/10.1007/978-3-319-48881-3_56

Advisor: Ketil Malde (ketil.malde@hi.no)

## Representation learning for object detection

While traditional classifiers work well with data that is labeled with disjoint classes and reasonably balanced class abundances, reality is often less clean. An alternative is to learn a vectors space embedding that reflects semantic relationships between objects, and deriving classes from this representation. This is especially useful for few-shot classification (ie. very few examples in the training data).

The task here is to extend a modern object detector (e.g. Yolo v8) to output an embedding of the identified object. Instead of a softmax classifier, we can learn the embedding either in a supervised manner (using annotations on frames) by attaching an ArcFace or other supervised metric learning head. Alternatively, the representation can be learned from tracked detections over time using e.g. a contrastive loss function to keep the representation for an object (approximately) constant over time. The performance of the resulting object detector will be measured on underwater videos, targeting species detection and/or indiviual recognition (re-ID).

Advisor: Ketil Malde (ketil.malde@hi.no)

## Time-domain object detection

Object detectors for video are normally trained on still frames, but it is evident (from human experience) that using time domain information is more effective. I.e., it can be hard to identify far-off or occluded objects in still images, but movement in time often reveals them.

Here we will extend a state of the art object detector (e.g. yolo v8) with time domain data. Instead of using a single frame as input, the model will be modified to take a set of frames surrounding the annotated frame as input. Performance will be compared to using single-frame detection.

Advisor: Ketil Malde (ketil.malde@hi.no)

## Large-scale visualization of acoustic data

The Institute of Marine Research has decades of acoustic data collected in various surveys. These data are in the process of being converted to data formats that can be processed and analyzed more easily using packages like Xarray and Dask.

The objective is to make these data more accessible to regular users by providing a visual front end. The user should be able to quickly zoom in and out, perform selection, export subsets, apply various filters and classifiers, and overlay annotations and other relevant auxiliary data.

Advisor: Ketil Malde (ketil.malde@hi.no)

## Learning acoustic target classification from simulation

Broadband echosounders emit a complex signal that spans a large frequency band. Different targets will reflect, absorb, and generate resonance at different amplitudes and frequencies, and it is therefore possible to classify targets at much higher resolution and accuracy than before. Due to the complexity of the received signals, deriving effective profiles that can be used to identify targets is difficult.

Here we will use simulated frequency spectra from geometric objects with various shapes, orientation, and other properties. We will train ML models to estimate (recover) the geometric and material properties of objects based on these spectra. The resulting model will be applied to read broadband data, and compared to traditional classification methods.

Advisor: Ketil Malde (ketil.malde@hi.no)

## Online learning in real-time systems

Build a model for the drilling process by using the Virtual simulator OpenLab (https://openlab.app/) for real-time data generation and online learning techniques. The student will also do a short survey of existing online learning techniques and learn how to cope with errors and delays in the data.

Advisor: Rodica Mihai

## Building a finite state automaton for the drilling process by using queries and counterexamples

Datasets will be generated by using the Virtual simulator OpenLab (https://openlab.app/). The student will study the datasets and decide upon a good setting to extract a finite state automaton for the drilling process. The student will also do a short survey of existing techniques for extracting finite state automata from process data. We present a novel algorithm that uses exact learning and abstraction to extract a deterministic finite automaton describing the state dynamics of a given trained RNN. We do this using Angluin's L*algorithm as a learner and the trained RNN as an oracle. Our technique efficiently extracts accurate automata from trained RNNs, even when the state vectors are large and require fine differentiation.arxiv.org

Advisor: Rodica Mihai

## Machine Learning for Drug Repositioning in Parkinson’s Disease

**Background**: Parkinson’s Disease (PD) is a major neurological condition with a complex etiology that tends to affect the elderly population. Understanding the risk factors associated with PD, including drug usage patterns across different demographics, can provide insights into its management and prevention. The Norwegian Prescribed Drug Registry (NorPD) provides comprehensive data on prescriptions dispensed from 2004, making it an excellent resource for such an analysis.

**Objective**: This project seeks to investigate how well machine learning techniques can predict PD risk, using the individual histories of drug usage along with demographic variables like gender and age.

**Methodology**:

- Exploratory Data Analysis and Data Preprocessing: Although the dataset is clean and structured, specific preprocessing steps will be required to tailor the data for the chosen methods.
- Predictive Modeling: Apply standard machine learning models such as Random Forest for handling large, imbalanced, sparse dataset, to find the best model or ensemble models to robust prediction. The predictive model will be employed to discern patterns in drug usage and demographic factors that correlate with PD risk.
- Feature Analysis: Conduct a detailed analysis to understand the importance of different features, such as specific drugs, gender, and age, in predicting PD risk and explore complex dependencies between features.
- Evaluation Metrics: Explore different metrics, such as F1-score and AUC-ROC to evaluate the performance of the predictive models.

**Expected Outcomes**: The project aims to study and develop predictive models that can accurately identify individuals at increased risk of developing PD based on their prescription history and demographic data.

**Ethical Considerations**: Data privacy and confidentiality will be strictly maintained by conducting all analyses on the SAFE server, following ethical guidelines for handling sensitive health data. The approval from regional ethics committee (REK) is already in place, as the project will be part of DRONE (https://www.uib.no/en/epistat/139849/drone-drug-repurposing-neurological-diseases).

**Project Benefits**.

- The student practices working with a huge and rich set of real data and working with experts from epidemiology group at MED faculty.
- Utilizing different ML methods in real data
- The possibility of publication if the results are promising.

**Advisors**: Asieh Abolpour Mofrad, Samaneh Abolpour Mofrad, Julia Romanowska, Jannicke Igland

## Exploring Graph Neural Networks for Analyzing Prescription Data to Predict Parkinson’s Disease Risk

**Background**: Parkinson’s Disease (PD) significantly impacts the elderly, necessitating advanced computational approaches to predict and understand its risk factors better. The Norwegian Prescribed Drug Registry (NorPD) provides comprehensive data on prescriptions dispensed from 2004, presents an excellent opportunity to employ graph neural networks (GNNs), especially to analyze the temporal dynamics of prescription data.

**Objective**. The project aims to investigate the effectiveness of GNNs in analyzing time-dependent prescription data, focusing on various graph structures to understand how drug interactions and patient demographics influence PD risk over time.

**Methodology**:

- Exploratory Data Analysis and Data Preprocessing: Prepare the prescription data for GNN analysis by investigating different structures to represent the data as a graph. This step is a challenging step; we must investigate what is the best structure for a graph based on the existing GNN and temporal GNN methods. For instance, one might assign a graph to each individual and consider classification approaches, or defining a graph for all participants, and investigating the GNN methods for clustering or predicting nodes and edges.

Incorporate demographic features, such as age, gender, and education, into the graph. Additionally, explore how to integrate time-dependent features to reflect the dynamic nature of the prescription data effectively.

- Graph Neural Network Implementation: Apply graph neural network models such as Graph Convolutional Networks (GCNs) or Graph Attention Networks (GATs) that can process temporal graph data, based on the structure of our defined graph.
- Feature Analysis: Perform an in-depth analysis of the learned embeddings and node features to identify significant patterns and influential factors related to increased, decreased PD risk.
- Evaluation Metrics: Explore different metrics to evaluate the performance of the predictive models.

**Expected Outcomes**:

The project aims to study how graph neural networks (GNNs) can be utilized to analyze complex, time-dependent prescription data.

**Ethical Considerations**. All analyses will adhere to strict privacy protocols by conducting research on the SAFE server, ensuring that all individual data remains confidential and secure in compliance with ethical healthcare data management practices. The approval from regional ethics committee (REK) is already in place, as the project will be part of DRONE (https://www.uib.no/en/epistat/139849/drone-drug-repurposing-neurological-diseases)

**Project Benefits**:

- The student practices working with a huge and rich set of real data and working with experts from epidemiology group at MED faculty.
- Get familiar with GNNs as advanced ML methods and utilize them in real data.
- The possibility of publication if the results are promising.

**Advisors**: Samaneh Abolpour Mofrad, Asieh Abolpour Mofrad, Julia Romanowska, Jannicke Igland

## Scaling Laws for Language Models in Generative AI

Large Language Models (LLM) power today's most prominent language technologies in Generative AI like ChatGPT, which, in turn, are changing the way that people access information and solve tasks of many kinds.

A recent interest on scaling laws for LLMs has shown trends on understanding how well they perform in terms of factors like the how much training data is used, how powerful the models are, or how much computational cost is allocated. (See, for example, Kaplan et al. - "Scaling Laws for Neural Language Models”, 2020.)

In this project, the task will consider to study scaling laws for different language models and with respect with one or multiple modeling factors.

Advisor: Dario Garigliotti

## Applications of causal inference methods to omics data

Many hard problems in machine learning are directly linked to causality [1]. The graphical causal inference framework developed by Judea Pearl can be traced back to pioneering work by Sewall Wright on path analysis in genetics and has inspired research in artificial intelligence (AI) [1].

The Michoel group has developed the open-source tool Findr [2] which provides efficient implementations of mediation and instrumental variable methods for applications to large sets of omics data (genomics, transcriptomics, etc.). Findr works well on a recent data set for yeast [3].

We encourage students to explore promising connections between the fiels of causal inference and machine learning. Feel free to contact us to discuss projects related to causal inference. Possible topics include: a) improving methods based on structural causal models, b) evaluating causal inference methods on data for model organisms, c) comparing methods based on causal models and neural network approaches.

References:

1. Schölkopf B, Causality for Machine Learning, arXiv (2019): https://arxiv.org/abs/1911.10500

2. Wang L and Michoel T. Efficient and accurate causal inference with hidden confounders from genome-transcriptome variation data. PLoS Computational Biology 13:e1005703 (2017). https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1005703

3. Ludl A and and Michoel T. Comparison between instrumental variable and mediation-based methods for reconstructing causal gene networks in yeast. arXiv:2010.07417 https://arxiv.org/abs/2010.07417

Advisors: Adriaan Ludl, Tom Michoel

## Space-Time Linkage of Fish Distribution to Environmental Conditions

Background

Conditions in the marine environment, such as, temperature and currents, influence the spatial distribution and migration patterns of marine species. Hence, understanding the link between environmental factors and fish behavior is crucial in predicting, e.g., how fish populations may respond to climate change. Deriving this link is challenging because it requires analysis of two types of datasets (i) large environmental (currents, temperature) datasets that vary in space and time, and (ii) sparse and sporadic spatial observations of fish populations.

Project goal

The primary goal of the project is to develop a methodology that helps predict how spatial distribution of two fish stocks (capelin and mackerel) change in response to variability in the physical marine environment (ocean currents and temperature). The information can also be used to optimize data collection by minimizing time spent in spatial sampling of the populations.

Approach

The project will focus on the use of machine learning and/or causal inference algorithms. As a first step, we use synthetic (fish and environmental) data from analytic models that couple the two data sources. Because the ‘truth’ is known, we can judge the efficiency and error margins of the methodologies. We then apply the methodologies to real world (empirical) observations.

Advisors: Tom Michoel, Sam Subbey.

## Towards precision medicine for cancer patient stratification

On average, a drug or a treatment is effective in only about half of patients who take it. This means patients need to try several until they find one that is effective at the cost of side effects associated with every treatment. The ultimate goal of precision medicine is to provide a treatment best suited for every individual. Sequencing technologies have now made genomics data available in abundance to be used towards this goal.

In this project we will specifically focus on cancer. Most cancer patients get a particular treatment based on the cancer type and the stage, though different individuals will react differently to a treatment. It is now well established that genetic mutations cause cancer growth and spreading and importantly, these mutations are different in individual patients. The aim of this project is use genomic data allow to better stratification of cancer patients, to predict the treatment most likely to work. Specifically, the project will use machine learning approach to integrate genomic data and build a classifier for stratification of cancer patients.

Advisor: Anagha Joshi

## Unraveling gene regulation from single cell data

Multi-cellularity is achieved by precise control of gene expression during development and differentiation and aberrations of this process leads to disease. A key regulatory process in gene regulation is at the transcriptional level where epigenetic and transcriptional regulators control the spatial and temporal expression of the target genes in response to environmental, developmental, and physiological cues obtained from a signalling cascade. The rapid advances in sequencing technology has now made it feasible to study this process by understanding the genomewide patterns of diverse epigenetic and transcription factors as well as at a single cell level.

Single cell RNA sequencing is highly important, particularly in cancer as it allows exploration of heterogenous tumor sample, obstructing therapeutic targeting which leads to poor survival. Despite huge clinical relevance and potential, analysis of single cell RNA-seq data is challenging. In this project, we will develop strategies to infer gene regulatory networks using network inference approaches (both supervised and un-supervised). It will be primarily tested on the single cell datasets in the context of cancer.

Advisor: Anagha Joshi

## Developing a Stress Granule Classifier

To carry out the multitude of functions 'expected' from a human cell, the cell employs a strategy of division of labour, whereby sub-cellular organelles carry out distinct functions. Thus we traditionally understand organelles as distinct units defined both functionally and physically with a distinct shape and size range. More recently a new class of organelles have been discovered that are assembled and dissolved on demand and are composed of liquid droplets or 'granules'. Granules show many properties characteristic of liquids, such as flow and wetting, but they can also assume many shapes and indeed also fluctuate in shape. One such liquid organelle is a stress granule (SG).

Stress granules are pro-survival organelles that assemble in response to cellular stress and important in cancer and neurodegenerative diseases like Alzheimer's. They are liquid or gel-like and can assume varying sizes and shapes depending on their cellular composition.

In a given experiment we are able to image the entire cell over a time series of 1000 frames; from which we extract a rough estimation of the size and shape of each granule. Our current method is susceptible to noise and a granule may be falsely rejected if the boundary is drawn poorly in a small majority of frames. Ideally, we would also like to identify potentially interesting features, such as voids, in the accepted granules.

We are interested in applying a machine learning approach to develop a descriptor for a 'classic' granule and furthermore classify them into different functional groups based on disease status of the cell. This method would be applied across thousands of granules imaged from control and disease cells. We are a multi-disciplinary group consisting of biologists, computational scientists and physicists.

Advisors: Sushma Grellscheid, Carl Jones

## Machine Learning based Hyperheuristic algorithm

Develop a Machine Learning based Hyper-heuristic algorithm to solve a pickup and delivery problem. A hyper-heuristic is a heuristics that choose heuristics automatically. Hyper-heuristic seeks to automate the process of selecting, combining, generating or adapting several simpler heuristics to efficiently solve computational search problems [Handbook of Metaheuristics]. There might be multiple heuristics for solving a problem. Heuristics have their own strength and weakness. In this project, we want to use machine-learning techniques to learn the strength and weakness of each heuristic while we are using them in an iterative search for finding high quality solutions and then use them intelligently for the rest of the search. Once a new information is gathered during the search the hyper-heuristic algorithm automatically adjusts the heuristics.

Advisor: Ahmad Hemmati

## Machine learning for solving satisfiability problems and applications in cryptanalysis

Advisor: Igor Semaev

## Hybrid modeling approaches for well drilling with Sintef

Several topics are available.

**Background**

"Flow models" are first-principles models simulating the flow, temperature and pressure in a well being drilled. Our project is exploring "hybrid approaches" where these models are combined with machine learning models that either learn from time series data from flow model runs or from real-world measurements during drilling. The goal is to better detect drilling problems such as hole cleaning, make more accurate predictions and correctly learn from and interpret real-word data.

The "surrogate model" refers to a ML model which learns to mimic the flow model by learning from the model inputs and outputs. Use cases for surrogate models include model predictions where speed is favoured over accuracy and exploration of parameter space.

**Surrogate models with active Learning **

While it is possible to produce a nearly unlimited amount of training data by running the flow model, the surrogate model may still perform poorly if it lacks training data in the part of the parameter space it operates in or if it "forgets" areas of the parameter space by being fed too much data from a narrow range of parameters.

The goal of this thesis is to build a surrogate model (with any architecture) for some restricted parameter range and implement an active learning approach where the ML requests more model runs from the flow model in the parts of the parameter space where it is needed the most. The end result should be a surrogate model that is quick and performs acceptably well over the whole defined parameter range.

**Surrogate models trained via adversarial learning **

How best to train surrogate models from runs of the flow model is an open question. This master thesis would use the adversarial learning approach to build a surrogate model which to its "adversary" becomes indistinguishable from the output of an actual flow model run.

**GPU-based Surrogate models for parameter search **

While CPU speed largely stalled 20 years ago in terms of working frequency on single cores, multi-core CPUs and especially GPUs took off and delivered increases in computational power by parallelizing computations.

Modern machine learning such as deep learning takes advantage this boom in computing power by running on GPUs.

The SINTEF flow models in contrast, are software programs that runs on a CPU and does not happen to utilize multi-core CPU functionality. The model runs advance time-step by time-step and each time step relies on the results from the previous time step. The flow models are therefore fundamentally sequential and not well suited to massive parallelization.

It is however of interest to run different model runs in parallel, to explore parameter spaces. The use cases for this includes model calibration, problem detection and hypothesis generation and testing.

The task of this thesis is to implement an ML-based surrogate model in such a way that many surrogate model outputs can be produced at the same time using a single GPU. This will likely entail some trade off with model size and maybe some coding tricks.

**Uncertainty estimates of hybrid predictions (Lots of room for creativity, might need to steer it more, needs good background literature)**

When using predictions from a ML model trained on time series data, it is useful to know if it's accurate or should be trusted. The student is challenged to develop hybrid approaches that incorporates estimates of uncertainty. Components could include reporting variance from ML ensembles trained on a diversity of time series data, implementation of conformal predictions, analysis of training data parameter ranges vs current input, etc. The output should be a "traffic light signal" roughly indicating the accuracy of the predictions.

**Transfer learning approaches **

We're assuming an ML model is to be used for time series prediction

It is possible to train an ML on a wide range of scenarios in the flow models, but we expect that to perform well, the model also needs to see model runs representative of the type of well and drilling operation it will be used in. In this thesis the student implements a transfer learning approach, where the model is trained on general model runs and fine-tuned on a most representative data set.

(Bonus1: implementing one-shot learning, Bonus2: Using real-world data in the fine-tuning stage)

**ML capable of reframing situations **

When a human oversees an operation like well drilling, she has a mental model of the situation and new data such as pressure readings from the well is interpreted in light of this model. This is referred to as "framing" and is the normal mode of work. However, when a problem occurs, it becomes harder to reconcile the data with the mental model. The human then goes into "reframing", building a new mental model that includes the ongoing problem. This can be seen as a process of hypothesis generation and testing.

A computer model however, lacks re-framing. A flow model will keep making predictions under the assumption of no problems and a separate alarm system will use the deviation between the model predictions and reality to raise an alarm. This is in a sense how all alarm systems work, but it means that the human must discard the computer model as a tool at the same time as she's handling a crisis.

The student is given access to a flow model and a surrogate model which can learn from model runs both with and without hole cleaning and is challenged to develop a hybrid approach where the ML+flow model continuously performs hypothesis generation and testing and is able to "switch" into predictions of a hole cleaning problem and different remediations of this.

Advisor: Philippe Nivlet at Sintef together with advisor from UiB

## Explainable AI at Equinor

In the project Machine Teaching for XAI (see https://xai.w.uib.no) a master thesis in collaboration between UiB and Equinor.

Advisor: One of Pekka Parviainen/Jan Arne Telle/Emmanuel Arrighi + Bjarte Johansen from Equinor.

## Explainable AI at Eviny

In the project Machine Teaching for XAI (see https://xai.w.uib.no) a master thesis in collaboration between UiB and Eviny.

Advisor: One of Pekka Parviainen/Jan Arne Telle/Emmanuel Arrighi + Kristian Flikka from Eviny.

## Own topic

If you want to suggest your own topic, please contact Pekka Parviainen, Fabio Massimo Zennaro or Nello Blaser.