Research areas
Below is a list of the predefined master's thesis projects available at Simula; the list can be sorted according to the research area of interest.
A Jira extension for Benefits Management in IT-projects
It is important for IT development projects to estimate and keep track of the benefit of the software they are producing. You will help develop tools for this purpose.
A system for planning and analyzing crisis management exercises
Crisis response exercises are held regularly, but often with unclear learning objectives, unplanned data collection and inferior analyses. You will contribute to developing an exercise management system for structured planning, execution and analysis of crisis response exercises.
Adaptive Tests for Memory Training using Asymmetric Search Point Location
Memory training exercises are known to have a positive effect on improving the memorability of humans. A memory training task can be as simple as trying to remember a password of increasing complexity or a number of varying length.
AI-Based Detection of Atrial Fibrillation from ECG Data for e-Health Applications
In this project, students will develop an AI model for detecting Atrial Fibrillation (AF) from electrocardiogram (ECG) recordings. The project emphasizes the development of machine learning (ML) models for use in e-Health and Clinical Decision Support (CDS) systems. Using publicly available or simulated ECG datasets, students will train and evaluate deep learning models (e.g., CNN, RNN, LSTM) to classify ECG signals as normal or AF-affected.
AI-based Video Analysis and Processing for Sports (multiple topics)
Interested in AI and video technology? In sports tech used by elite leagues and live broadcasters? If either, read on ...
Augmentation and Generation of Biomedical Signals for Privacy- Preserving AI
The students will explore and implement data augmentation techniques and generative models for biomedical signals. These methods may include traditional signal augmentation (e.g., noise addition, scaling, time-warping), as well as deep learning–based generative techniques such as TimeGAN, 1D-VAEs, or simple CNN-based generative pipelines.
Automatic detection of abnormal video events in sport videos
Use machine learning and video processing techniques to automatically find events in sports videos. For example, in football, some "easy" ones are goals, but others like tackles are harder.
Autonomous Self-healing Software Systems
Investigate, develop and evaluate data-driven techniques and prototypes that help software engineers build software systems that are autonomously self-healing. These are systems that can understand when they are not operating correctly and, without human intervention, make the necessary adjustments to restore themselves to normal operation.
Benchmarking Modern AI Hardware for Natural Language Processing
A new generation GPUs and AI accelerators such as IPUs promise massive speedups for NLP. Do they work out in practice?
Benchmarking Partitioning tools for Supercomputers
How can we make best use of partitioning software and distribute workloads among compute nodes in a supercomputer?
Benefit points - what will it take for IT professionals to use them
Just as story points are estimates of the cost of a piece of IT functionality, benefit points are estimates of the intended benefit (value for stakeholders) of that piece of functionality. It seems obvious that one should use benefit points in addition to story points, but benefit points (or similar metrics) are not in widespread use.
Bio-inspired active sensing using reinforcement learning
Explore how reinforcement learning can drive sensory-motor integration in distributed systems inspired by the whisker barrel cortex, training on images with movable sensors, and potential 3D object recognition.
Building and evaluating a web-based tool for software benefits estimation and management
To help IT professionals optimize the value for stakeholders, you will develop a set of easy-to-use tool for estimating and monitoring how much potential benefit a system under development will produce. You will design the tool based on recent theoretical results, and you will test and evaluate the tool with actual software professionals.
Chain of Thought vs Direct Answers: Cost, Latency, and Quality at Scale
Test when hidden or minimal reasoning beats verbose chains.
Classification confidence visualization of artificial neural networks with adversarial robustness
Adversarial attacks can easily fool artificial neural networks. This project aims to understand these attacks and their defenses by visualizing their classification confidence.
Co-production and co-destruction patterns in systems design, development and use
An IT system is meant for creating benefit for both service providers and service consumers. Often, disbenefit is created instead. You will study the phenomena of co-production of benefit and co-destruction into disbenefit in public service IT systems.
Combining metrics of energy efficiency and complexity of AI algorithms
Nowadays the high performance of AI algorithms developed together with an ever increasing complexity and huge energy cost. This makes nowadays AI less and less sustainable and future AI research should account for these trade-offs.
Competitive Influence Maximization: Countering Disinformation with Algorithms
How can we best fight influence campaigns of trolls and bots? Competitive influence maximization uses graph algorithms and the structure of social networks to defeat the attackers.
CPU free Programming: When the GPU takes the Lead
New tools can remove inefficiencies in GPU computing, but can we turn that into real performance gains?
Detecting DDoS Attacks in Programmable Data Planes
Building a machine learning model using data plane programming language such as P4 that detect network security attacks at line rate with high accuracy
Development and Testing of Self-Driving Cars
Implementing software tools for designing, developing, and testing of self-driving cars.
Development of a Mobile Application for Automated Scrotal Circumference Measurement in Beef Bulls Using iPhone LiDAR and Deep Learning
Mobile app development for a real world biological use case. Involves machine learning, image processing and app development as part of a research project.
DopplerLog: A Performance Measurement Tool for Composable PCIe Infrastructures
e-Health Application for Detecting Chronic Cardiovascular Diseases: Ischemic Heart Disease
In this project, students will develop an AI model for detecting ischemic heart disease (IHD) from medical images, potentially including both cardiac magnetics resonance images (CMR) and contrast computed tomography (CCT) of the chest, heart, and/or coronary vessels. The exact nature of the imaging to be included will be determined at project start.
Enhancing Patch Validation in Automated Program Repair through Large Language Models
Investigate and develop methods to confirm the correctness of patches generated by APR systems, leveraging the capabilities of large language models (LLMs).
Evolving neural networks for optimal foraging
If a neural network produces an output that switches at random, can evolution be used to shape this noise? Can this help explain the random movement patterns of animals foraging for food?
Explainability of time series missing data techniques
Dive into an experimental study to discover how filling missing values for time series data affects the explainability of a downstream classification task. Potential applications are health care, sport analysis or lifelogging (the application of your choice).
Explainable Reinforcement Learning
In recent years, artificial intelligence has made significant strides in various fields, reshaping the landscape of technology and innovation. One of the key factors driving this progress is the emergence of reinforcement learning (RL) which enables autonomous agents to make decisions and adapt to their surroundings. RL has been very successful in games and other applications. In general, an RL agent aims to learn a near-optimal policy to achieve a fixed objective by taking action and receiving feedback through rewards and observations from the environment. A neural network (NN) commonly represents the policy that, given the observation of the environment state as input, yields values that indicate which action to choose.
Exploring Multidomain Applications of Large Language Models in Software Engineering
This project meets the demand for enhanced approaches by harnessing LLMs to elevate software engineering practices in specific research domains.
Extracting insights from multiple metabolomics data sets through data fusion
The project focuses on joint analysis of NMR (Nuclear Magnetic Resonance) spectroscopy measurements of plasma and urine samples as well as faecal metabolome data using interpretable multimodal data mining.
Generative machine learning for precision medicine
The accuracy of machine learning models used in clinical decision making has a direct impact on a patient's chances of recovery. Missing data pose a challenge and generative models can assist overcome it.
GPU Performance Optimization: Perfect Matchings and the Wasserstein distance
The Wasserstein distance is a key metric in machine learning, but it is hard to compute optimally on GPUs. By using approximation algorithms, we can unlock the full power of the GPU.
High Performance Computing in the Mojo Programing Language
Mojo is a programming language that promises to combine the productivity of Python with the performance of C++. Can it fulfill this promise? Let’s find out!
Image resolution vs sperm analysis using AI
Uncover the impact of image resolution on sperm analysis using machine learning! Dive into an experimental study that explores how video quality influences the accuracy of AI in predicting sperm motility and morphology.
Impact of contrastive learning methods for training foundation models in ECG analysis
Dive into an experimental study to discover the best methods for applying contrastive learning to ECG data, including algorithm selection, data pairing techniques, and optimal data formats for deep learning.
Investigating Smart NICs in PCI Express Networks
Investigate the possibilities of using a Smart NIC to do in-network processing in a PCI Express network.
Knowledge-guided machine learning for interpretable pattern discovery
The goal of the project is to develop unsupervised machine learning methods that will guide real data analysis with mechanistic models and reveal interpretable patterns to extract insights from complex data.
Large Language Models Adaptation for Cyber-Physical System Testing
Dive into the challenge of testing Cyber-Physical Systems (CPSs) by optimizing and leveraging the potential of Large Language Models (LLMs).
LLM-Driven Testing: Assessing Large Language Models in Cancer Registry Applications
This research project seeks to transform cancer registry testing by harnessing the power of Large Language Models (LLMs) like ChatGPT, offering automated, generative testing methods to detect anomalies, create test cases, and enhance data quality.
Modeling Hippocampal Spatial Navigation and Memory using Brain-Inspired AI
Delve into the fascinating intersection of neuroscience and AI by creating brain-inspired AI models of the hippocampus to investigate how we navigate our world and form new memories.
Modeling the mechanics of the heart
The heart is the organ responsible for pumping blood around in your body. The heart consist of a tissue known as myocardium which is known to be a anisotropic, nonlinear, visco-elastic and nearly incompressible material. In order to create realistic models of the mechanics of the heart we would therefore need to incorporate these effects as well as appropriate boundary conditions.
Modelling and simulating the brain's waterscape
Modelling the evolution of the New York Stock Exchange
This project handles with the 10-min NY Stock Exchange data between 2011 and 2014 and aims at processing the data and publish it for research purposes.
Modelling the Heart's Hidden Hero: The Right Ventricle
Continuing the quest to unlock the mysteries of right ventricular function through machine learning and computational modeling
Monitoring 2d environments with eye-tracking based agents
What kind of environments are our visual search strategies adapted to? Can they be transplanted to autonomous vehicles?
Multimodal-LLMs for Explanations in Automated Driving
Evaluate LLMs for enhancing trustworthiness in automated driving through identification of relevant objects and anticipation of their future behaviour and explanation of the actions taken by the car.
Neuro-Symbolic Models for Scene Understanding in Automated Driving
Develop a neuro-symbolic pipeline that combines scene graphs and machine learning to identify relevant objects for automated driving manoeuvres.
Optimizing Network Efficiency in Modern Supercomputers
How can we distribute workloads among compute nodes to make best use of the high speed networks in supercomputers?
Parallel implementation of graph neural networks
Graph Neural Networks are a powerful machine learning technique for unstructured data. How can we make them scale to supercomputers?
PCI Express support for BeeGFS
In this master, you will add RDMA functionality to the global shared filesystem BeeGFS™️ (ThinkParQ) when running over PCIe NTB interconnects from Dolphin Interconnect Solution.
Performance monitoring and diagnosis in mobile communication networks
Build an automated system based on machine learning to diagnose fault conditions in mobile communication networks.
Predicting Next Purchase Day and Order Volume for Customers in a Supply Chain
TINE SA serves over 20,000 business customers who place orders directly with the company. Each customer exhibits unique purchasing behaviors influenced by various factors, including seasonality and product shelf-life constraints. These behaviors are further shaped by individual seasonality patterns (such as stable or variable seasonal dates, geographically specific holidays, and annual events) and distinct warehouse management strategies (such as stockpiling or maintaining a consistent order policy).
Prediction of dry eye disease using metabolomics data
Development of methods to improve metabolomics data handling for predicting dry eye disease.
Programing AMD GPUs for High performance Computing
AMD GPUs are very powerful, but how can we use their full potential?
Quantifying individual differences in eye-gaze dynamics
We know some people, such as those with ADHD and autism, tend to gaze at images differently. But at the individual level, could everyone have their “own” way of looking? Can we use this to improve diagnoses and better understand visual search?
Quantum Software engineering and its applications
Quantum computing is at the edge of a technological revolution, offering you the chance to be part of the innovation that will solve problems once thought impossible. Jump in now and help shape the future of this groundbreaking field!
Real-world optimization and machine learning on quantum computers
Up for quantum-powered solutions to real-world optimization or machine learning?
Securing the AI pipeline: Privacy and security metrics for trustworthy and transparent AI
Nowadays transparency is an important aspect of the Trustworthiness AI algorithms should have. However, making AI more transparent raises issues concerning privacy and security. How to best balance the transparency an AI algorithm should have for users with an also needed security and privacy of the data it accesses/uses?
Sperm-net for ML research
Contribute to the future of AI-driven sperm analysis by curating, organizing, and testing open-access datasets.
Sport news classification
Automatic classification of sport news into different categories using state of the art Natural Language Processing (NLP). The project is building upon an existing dataset of Norwegian soccer news.
Synthetic Cardiac MRI Image Generation using Deep Generative Models
In this project, students will explore and evaluate deep learning models for generating synthetic CMRI images using limited real samples. They will implement and compare generative approaches such as Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), and Diffusion Models. The project may also explore few-shot learning strategies and inpainting techniques to enhance synthetic image quality and diversity.
Synthetic Medical Tabular Data Generation using Deep Generative Models
In this project, students will focus on the generation of synthetic tabular medical data using deep learning methods. They will design and evaluate models such as Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), or Diffusion Models to create data resembling structured patient records (e.g., demographics, lab test values, and diagnoses).
Topic Classification of Scientific Papers
Millions of scientific papers are published very year. Can we use AI to understand their contents?
Towards Reliable Solutions from Artificial Intelligence: From Learning to Decision-Making
How can we use mathematical optimization, uncertainty quantification, and asymptotic statistics to develop a computational framework that takes us from data and training statistical models to actually making robust, risk-averse decisions based on the predictions of these models?
Uncertainty quantification of missing data
Dive into an experimental study to discover the best methods for quantifying uncertainty when filling in missing values for time series data and how this affects the uncertainty quantification in downstream classification tasks. Potential applications are health care, sport analysis, or lifelogging (the application of your choice).
Uncovering extreme climate events with machine learning
One of the most worrisome consequences of climate change for modern societies is the occurrence of extreme events, in particular heatwaves, wildfires and droughts. Extreme events impact not only our environment, but also our economy and health, so our society in general.
Unlock the Future of Medical AI with the Kvasir-VQA Dataset
We aim to benchmark the Kvasir-VQA dataset across various cutting-edge tasks. As a student, you have the flexibility to choose one or multiple tasks that align with your interests and focus your research efforts accordingly.
Unlocking the Potential of Digital Twins for Software Systems
Crafting digital twins for diverse software systems.
Use of a Bayesian Approach for Generating Statistically Representative Sky Signals
Apply advanced Bayesian techniques to synthesize statistically representative sky signals. This project focuses on implementing Gibbs sampling for astrophysical data, offering a rigorous approach to overcoming optimization challenges in modern cosmological analysis.
Visualizing Benefits: A Systematic Approach to Benefits Management via a Digital Platform
We are embarking on an exciting project to build a digital platform that tackles one of the most challenging areas for project managers and organizations - Benefits Management. Despite its increasing importance in recent years, many organizations find themselves navigating the complex world of benefits management with no clear direction. This is where you come in. By joining our project, you will be at the forefront of creating a solution that demystifies benefits management. You will help provide clear and streamlined processes for planning, monitoring, and reporting, ultimately leading organizations to success in their benefits management endeavors.
What can we do with a quantum computer with only two or three q-bits?
This is in fact a set of possible projects in the fields of AI and quantum computation. We will use the first quantum computers of Norway recently bought by the Department of Computer Science and the OsloMet AI Lab.