As a machine learning engineer, I'm deeply interested in the fields of machine learning and artificial intelligence, particularly in Meta-Learning and Reinforcement Learning. My goal is to build agents that can learn from new experiences and efficiently adapt to new problems or environments.
Research
Regularized Meta-Learning for Neural Architecture Search
Rob van Gastel, Joaquin Vanschoren
AutoML-Conf Late-Breaking Workshop 2022
paper / code
We apply regularization techniques to the inner-loop neural architecture search to improve meta-learning, adapting to new tasks more quickly.
Projects
Meta-Reinforcement Learning Algorithms
Currently working on implementing Meta-Reinforcement Learning algorithms designed to quickly adapt policies to new, related tasks within a few episodes.
Finetuning Transferable Vision Transformer Weights
Finetuning the encoder weights of the self-supervised learning method DINOv2 using a simple 1x1 convolution encoder and Low-Rank Adaptation (LoRA) allows for adaptation to the Pascal VOC and ADE20k datasets within a few epochs.
Discrete Soft Actor-Critic with Memory-Based Networks
Implemented the discrete Soft-Actor Critic (SAC) algorithm with RNN-conditioned policy and value functions. This conditioning is used to address challenges in partially observable environments.
A native iOS interval timer app to help people time their workouts in collaboration with Max Wammes and Mehmet Bakirci.