MSc Machine Learning
The master’s programme in Machine Learning gives you a foundation in mathematics, statistics and algorithms for understanding and developing systems that learn from data. You combine theory with practice and can explore deep learning and generative AI, computer vision, language technology and robotics. You learn to frame problems for machine learning and design, implement and critically evaluate solutions, preparing you for advanced roles in industry and research or for doctoral studies.
Machine Learning at KTH
Machine learning is transforming how complex problems are approached across science, engineering and industry. Recent advances in generative AI, large language models and agentic AI demonstrate both the pace of progress and the growing reach of the field. In this programme, you study important emerging technologies while developing a deep understanding of the principles and methods that underpin modern machine learning. This foundation enables you to critically evaluate and effectively apply current techniques, develop new solutions, and continue adapting as the field evolves.
Building on this foundation, the programme spans probabilistic modelling, optimisation and statistical learning, as well as neural networks, representation learning, deep learning and artificial intelligence. Across different application domains, you examine how the properties of the data and task influence model design, training and evaluation. Depending on your course choices, you can study deep generative modelling in greater depth, including autoregressive models and large language models, diffusion models and other generative approaches. Other options let you investigate robustness, explainability, fairness and privacy in machine learning systems.
Teaching moves between mathematical reasoning, implementation and empirical investigation. You derive and prove selected results, implement methods from mathematical or algorithmic descriptions, and train, tune and evaluate models on data. Laboratory work, programming assignments and projects are combined with seminars and written and oral work, developing your ability to analyse scientific literature, design sound experiments and explain what the evidence supports. The aim is that you can work independently on problems for which neither the appropriate method nor the answer is given in advance.
The programme draws on KTH research environments spanning machine learning, reinforcement learning, computer vision and robotics, speech and music technology, computational science and the theoretical foundations of data science. This gives you contact with both fundamental research and applications, as well as opportunities to engage with current research questions in courses and projects.
This is a two-year programme (120 ECTS credits) given in English. Graduates are awarded the degree of Master of Science. The programme is given mainly at KTH Campus in Stockholm by the School of Electrical Engineering and Computer Science (at KTH).
Programme structure and progression
During the first year, you build a foundation in machine learning through mandatory courses in machine learning, artificial intelligence and research methodology. These courses provide the basis for more advanced studies later in the programme.
From the second semester onwards, you shape your own profile through elective courses from two broad areas: application domains that use machine learning and theoretical machine learning. Depending on your interests, you can deepen your knowledge in areas such as computer vision, information retrieval, speech and language processing, computational biology, robotics, statistics, optimisation, deep learning and generative AI.
The programme concludes with a degree project carried out during the final semester in an academic or industrial environment in Sweden or abroad. In the past, students from the programme have completed projects at companies such as Saab, Elekta, Flir, Ericsson, Tobii, Spotify, Thales, and AstraZeneca.
Courses in the programme
The courses in the programme cover topics such as machine learning, deep learning, statistical modelling, artificial intelligence, computer vision, speech technology, information retrieval, and optimisation.
Courses in the master's programme in Machine Learning
Future and career
Graduates work as machine learning engineers, AI engineers, deep learning engineers, computer vision engineers, MLOps engineers, data scientists, data analysts, data engineers, software developers, software engineers, systems engineers, quantitative analysts, researchers and technical consultants. You can find them in these roles at technology companies and start-ups, as well as in sectors such as telecommunications, healthcare and pharmaceuticals, manufacturing, retail, media and entertainment, finance and consulting. Graduates from the programme have pursued careers at companies such as Ericsson, Google, Spotify, Netlight, Klarna, Modulai, H&M, Scania, AstraZeneca, Logitech, Dice and McKinsey, in Sweden and internationally.
The programme prepares you to work across the full machine learning development process, from preparing and analysing data and designing experiments to training and evaluating models and integrating them into larger software systems. This work can involve building recommendation and personalisation systems for digital services, analysing medical or industrial images, developing computer vision, language technologies and generative AI, enabling robots to perceive and act, or using predictive models in manufacturing, finance and decision support. Graduates have, for example, developed machine learning models at Klarna to understand online payment behaviour and personalise the purchase experience, and applied AI to advanced antenna systems and radio networks at Ericsson. The programme also prepares you to address the ethical and professional responsibilities that arise when AI systems are put into use. These include assessing their reliability, robustness and fairness, considering privacy and transparency, and
The programme provides a strong foundation for research and development roles in industry and for doctoral studies. Graduates have progressed to positions as senior specialists, technical leads, researchers and engineering managers.
Sustainable development
Graduates from KTH have the knowledge and tools for moving society in a more sustainable direction, as sustainable development is an integral part of all programmes. The three key sustainable development goals addressed by the master's programme in Machine Learning are:
Machine learning and generative AI can support advances in healthcare, renewable energy, sustainable cities and agriculture. Their benefits may not be equally distributed, however, and their use can have unintended consequences. As these technologies become more widely used, they also raise questions about fairness, privacy, transparency and reliability, as well as how automation changes work and professional roles.
In the mandatory parts of the programme, you examine what machine learning can and cannot predict, how models and results should be evaluated and communicated, and the responsibilities involved in using data and AI systems. Depending on your course choices, you can explore these questions in greater depth through topics such as bias mitigation, interpretability and explainability, uncertainty, robustness, privacy-preserving methods, and the ethical and societal aspects of generative AI.
The programme connects these perspectives to the UN Sustainable Development Goals through examples such as improving the operation of wind and solar farms, supporting medical diagnosis, personalised treatment and health interventions, and making farming practices more efficient through precision technologies.
The degree project also gives you the opportunity to investigate how machine learning can contribute to sustainable development. Previous degree projects have addressed:
-
Good Health and Well-being, in collaboration with medical technology companies such as Elekta and RaySearch;
-
Sustainable Cities and Communities, through the automatic monitoring of satellite imagery at KTH’s Division of Geoinformatics;
-
Peace, Justice and Strong Institutions, in collaboration with the Stockholm International Peace Research Institute (SIPRI).
Faculty and research
The programme is taught by faculty from several research environments within KTH’s School of Electrical Engineering and Computer Science, with a particularly strong connection to the Department of Robotics, Perception and Learning (RPL) . Researchers at RPL work at the intersection of machine learning, computer vision and robotics. Their research covers areas such as active perception, representation learning, mobile robotics, grasping and manipulation, planning and decision-making, and social robotics.
Research at RPL explores how intelligent systems can perceive and interpret their surroundings, learn from data, plan actions and interact with people. Applications range from industrial and service robotics to search and rescue, medical applications and assistive technologies for older people. Through courses and projects taught by active researchers, you encounter both fundamental questions in machine learning and the challenges involved in applying it in autonomous systems operating in complex environments.
KTH is also one of the partner universities in the Wallenberg AI, Autonomous Systems and Software Program (WASP) , Sweden’s largest individual research programme. WASP supports strategically motivated basic research, education and faculty recruitment in artificial intelligence, autonomous systems and software. Through its research programme, strategic recruitment, national graduate school and collaboration with industry, WASP strengthens the wider research environment in which the programme is taught.
