Computer Science

Permanent URI for this collectionhttps://hdl.handle.net/10679/43

Browse

Browsing by Rights "Attribution-NonCommercial 4.0 International"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    ArticlePublicationOpen Access
    Emotion as an emergent phenomenon of the neurocomputational energy regulation mechanism of a cognitive agent in a decision-making task
    (Sage, 2021-02) Kırtay, M.; Vannucci, L.; Albanese, U.; Laschi, C.; Öztop, Erhan; Falotico, E.; Computer Science; ÖZTOP, Erhan
    Biological agents need to complete perception-action cycles to perform various cognitive and biological tasks such as maximizing their wellbeing and their chances of genetic continuation. However, the processes performed in these cycles come at a cost. Such costs force the agent to evaluate a tradeoff between the optimality of the decision making and the time and computational effort required to make it. Several cognitive mechanisms that play critical roles in managing this tradeoff have been identified. These mechanisms include adaptation, learning, memory, attention, and planning. One of the often overlooked outcomes of these cognitive mechanisms, in spite of the critical effect that they may have on the perception-action cycle of organisms, is “emotion.” In this study, we hold that emotion can be considered as an emergent phenomenon of a plausible neurocomputational energy regulation mechanism, which generates an internal reward signal to minimize the neural energy consumption of a sequence of actions (decisions), where each action triggers a visual memory recall process. To realize an optimal action selection over a sequence of actions in a visual recalling task, we adopted a model-free reinforcement learning framework, in which the reward signal—that is, the cost—was based on the iteration steps of the convergence state of an associative memory network. The proposed mechanism has been implemented in simulation and on a robotic platform: the iCub humanoid robot. The results show that the computational energy regulation mechanism enables the agent to modulate its behavior to minimize the required neurocomputational energy in performing the visual recalling task.
  • Thumbnail Image
    Conference ObjectPublicationOpen Access
    A learning-based dependency to constituency conversion algorithm for the turkish language
    (European Language Resources Association (ELRA), 2022) Marşan, B.; Yıldız, O. K.; Kuzgun, A.; Yenice, A; Cesur, N.; Yenice, A. B.; Sanıyar, E.; Kuyrukçu, O.; Arıcan, B. N.; Yıldız, Olcay Taner; Computer Science; YILDIZ, Olcay Taner
    This study aims to create the very first dependency-to-constituency conversion algorithm optimised for Turkish language. For this purpose, a state-of-the-art morphologic analyser (Yıldız et al., 2019) and a feature-based machine learning model was used. In order to enhance the performance of the conversion algorithm, bootstrap aggregating meta-algorithm was integrated. While creating the conversation algorithm, typological properties of Turkish were carefully considered. A comprehensive and manually annotated UD-style dependency treebank was the input, and constituency trees were the output of the conversion algorithm. A team of linguists manually annotated a set of constituency trees. These manually annotated trees were used as the gold standard to assess the performance of the algorithm. The conversion process yielded more than 8000 constituency trees whose UD-style dependency trees are also available on GitHub. In addition to its contribution to Turkish treebank resources, this study also offers a viable and easy-to-implement conversion algorithm that can be used to generate new constituency treebanks and training data for NLP resources like constituency parsers.