Publications by authors named "Kristjan Korjus"
Evolution of cooperation and competition can appear when multiple adaptive agents share a biological, social, or technological niche. In the present work we study how cooperation and competition emerge between autonomous agents that learn by reinforcement while using only their raw visual input as the state representation. In particular, we extend the Deep Q-Learning framework to multiagent environments to investigate the interaction between two learning agents in the well-known video game Pong.
View Article and Find Full Text PDFArticle Synopsis
- Supervised machine learning needs to divide data into training, validation, and test sets, which complicates finding the best model parameters when data is limited.
- A new method called "Cross-validation and cross-testing" allows researchers to reuse test data without introducing bias, thus improving overall classifier performance evaluation.
- This approach has shown better results in identifying significant findings while keeping parameters interpretable, making it a valuable addition to existing machine learning techniques, especially for models where parameter interpretation is crucial.
In the present study we asked whether it is possible to decode personality traits from resting state EEG data. EEG was recorded from a large sample of subjects (n = 289) who had answered questionnaires measuring personality trait scores of the five dimensions as well as the 10 subordinate aspects of the Big Five. Machine learning algorithms were used to build a classifier to predict each personality trait from power spectra of the resting state EEG data.
View Article and Find Full Text PDF