Institutional Repository of Vilnius University
Canonical quantum recurrent neural network validation /
Abstract
dc:descriptionFollowing the work done in "Quantum Recurrent Neural Networks for Sequential Learning" by Yanan Li et al., we aim to validate the proposed Quantum Recurrent Neural Network models. To achieve this, we implement and train the models as faithfully as possible to the original paper. Our training and performance measurements utilize financial data from three stocks with varying price ranges. This study provides a concise overview of Recurrent Neural Networks and Quantum Recurrent Neural Networks. Additionally, we reiterate the known details about the proposed QRNNs and outline potentially crucial aspects that may be missing. Once the concepts used in the original work are introduced, we describe our implementation of the model variants, data handling methods, and the training routine. Subsequently, we conduct several experiments using financial data and identify probable underlying causes for the variations in performance. It is important to note that the models cannot be fully validated due to several key parameters and process details, such as perturbation, data handling, and training epochs, being obscured in the original work. Nevertheless, if these issues are resolved, the proposed models have the potential to fulfill the role of canonical QRNNs.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Pupšys, Aurimas,
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD192847202&prefLang=en_US