You've come to Dirk's github profile.
- I’m currently working in the technology department of an oil major.
- I also teach at UiO (where I'm an adjunct aossciate professor) how to make data-driven decision using ML.
- Fun fact: I also gave one of the first ever lectures in data science at UiO.
- Lecture materials
- IN-STK5000 – Responsible Data Science (2024)
- IN-STK5000 – Adaptive methods for data-based decision making (2023)
- IN-STK5000 – Adaptive methods for data-based decision making (2022)
- IN-STK5000 – Adaptive methods for data-based decision making (2021)
- IN-STK5000 – Adaptive methods for data-based decision making (2020)
- IN-STK5000 – Adaptive methods for data-based decision making (2019)
- Big Data Analytics Summer School 2017 at Oslo Met
- STK-INF4000 – Selected Topics in Data Science (2017)
- In line with my passion for teaching I'm regularly engaged in public speaking.👨🏫
- You can find me here:
Said's thesis, "Modeling Electricity Consumption with Mobility Data: Identifying Opportunities for Localized Energy Markets," investigates the potential of using public bike-sharing data to improve electricity consumption forecasting and grid management in Oslo, Norway.
This thesis explores the application of deep reinforcement learning, specifically deep policy gradient methods, for creating a trading agent that operates in volatile commodity markets.
This thesis focuses on using unsupervised machine learning to detect anomalies in large, complex cybersecurity datasets, specifically focusing on firewall logs. The research argues that traditional cybersecurity methods struggle to detect new and subtle threats, necessitating more advanced approaches.
The study uses various machine learning techniques, including logistic regression and random forests, to analyze features extracted from different aspects of Twitter accounts: