I am a driven and enthusiastic individual with a strong academic background in Artificial Intelligence and Data Science. Recently, I completed my bachelor's thesis as part of my studies at the Technical University of Denmark (DTU), where I specialized in variational inference and Bayesian Machine Learning. Currently, I am studying for a masters degree in Mathematical Modeling & Computation at DTU where I specialize in Machine Learning and Data Science. Besides my studies, I work as a software developer at a startup company that leverages AI to bridge gaps in the legal-tech sector.
My education has equipped me with a solid toolbox of skills, including:
- Machine Learning
- Deep Learning
- Data Mining
- Statistical Evaluation
I'm proficient in Python, and also experienced in JavaScript, PHP, SQL, and R. With a keen interest in cutting-edge AI technologies, I am well-prepared to contribute to shaping our future.
- Institution: Technical University of Denmark
Description: This project, part of my Bachelor's Thesis, explores the application of low-dimensional audio feature representations in a latent space for similarity-based search and retrieval. It investigates two methods: supervised transfer learning using YAMNet and unsupervised learning using a Variational Autoencoder (VAE). The project is evaluated on the ESC-50 dataset for environmental sound classification. While YAMNet significantly outperforms VAE, this research sheds light on the possibilities and challenges of using feature representation vectors for similarity search.
- Institution: Technical University of Denmark
Description: In response to the proliferation of fake news online, this project aims to develop a news article classification system using the Huggingface Transformer framework. By leveraging pre-trained models and enhancing classification performance, the objective is to distinguish between real and fake news articles. Utilizing NLP models such as RoBERTa, the project addresses the critical need for reliable tools to identify misleading content in the digital age.
- Institution: Technical University of Denmark
Description: This project, part of my study line project, focuses on diagnosing epilepsy through the analysis of electroencephalogram (EEG) signals. By preprocessing and handling EEG signals from Temple University Hospital, the project aims to classify electrode pop artifacts and build robust classifier models. Despite encountering challenges such as unbalanced datasets and computational constraints, the project highlights the potential for precise classification through experimentation with more complex models.
- Institution: Copenhagen MedTech
Description: I served as contributor and technical support for the hackathon described in this repo. The hackathon challenges participants to leverage artificial intelligence techniques to enhance EEG analysis. BrainCapture utilizes an advanced transformer model to map segments of EEG data into a latent, high-dimensional space, encapsulating pertinent information within the EEG signals. The task of the participants was to create a cloud pipeline for analyzing these latent representations.
In addition to my academic achievements, I have valuable professional experience:
- Student Assistant and IT Administrator at ProInvent A/S
- Software Developer at Aumento Advokatfirma/Legid.ai ApS
At Aumento Advokatfirma/Legid.ai ApS, I contribute to full-stack development of legal tech solutions using languages such as PHP, HTML, JavaScript, and SQL.
- Finalist in the Raymond Ideas Challenge for my project on intelligent solar power control using machine learning.
- Certifications from Science Talent Academy.
- Actively engaged in fostering an inclusive and supportive study life, both as a member of the study council and by teaching a voluntary AI course for women in collaboration with Microsoft Denmark.
I'm always eager to take on new challenges and collaborate with talented individuals. Connect with me on LinkedIn to explore opportunities in AI, data science, and software development.
Feel free to reach out! Let's build together. đâ¨