Method and system for target based hyper-parameter tuning Poorya ZareMoodi et al., 2021.
Learning to multi-task learn for better neural machine translation Poorya ZareMoodi, Gholamreza Haffari. arXiv, 2020.
Adaptively Scheduled Multitask Learning: The Case of Low-Resource Neural Machine Translation. Poorya ZareMoodi, Gholamreza Haffari. In Proceedings of 3rd WNGT, co-located with EMNLP, 2019.
Incorporating Syntactic Uncertainty in Neural Machine Translation with a Forest-to-Sequence Model. Poorya ZareMoodi, Gholamreza Haffari. In Proceedings of COLING, 2018.
Adaptive Knowledge Sharing in Multi-Task Learning: Improving Low-Resource Neural Machine Translation. Poorya ZareMoodi, Wray Buntine, Gholamreza Haffari. In Proceedings of ACL, 2018.
Neural Machine Translation for Bilingually Scarce Scenarios: A Deep Multi-task Learning Approach. Poorya ZareMoodi, Gholamreza Haffari. In Proceedings of NAACL, 2018.
Concept-evolution detection in non-stationary data streams: a fuzzy clustering approach. Poorya ZareMoodi, Sajjad Kamali Siahroudi, Hamid Beigy. Knowledge and Information Systems, 2018.
Detection of evolving concepts in non-stationary data streams: A multiple kernel learning approach. Sajjad Kamali Siahroudi, Poorya Zare Moodi, Hamid Beigy. Expert Systems With Applications, 2018.
A Support Vector Based Approach For Classification Beyond the Learned Label Space in Data Streams. Poorya ZareMoodi, Sajjad Kamali Siahroudi, Hamid Beigy. In Proceedings of ACM SAC, 2016.
A Novel Concept Drift Detection Method in Data Streams Using Ensemble Classifier. Mahdieh Dehghan, Hamid Beigy, Poorya ZareMoodi. Intelligent Data Analysis, 2016.
Novel Class Detection in Data Streams Using Local Patterns and Neighborhood Graph. Poorya ZareMoodi, Hamid Beigy, Sajjad Kamali Siahroudi. Neurocomputing, 2015.
My research interests lie in the fields of Deep Learning learning and Natural Language Processing. My PhD research was about improving Neural Machine Translation (NMT) in bilingually scarce scenarios. Like most deep learning approaches, NMT is notorious for its dependence on large amounts of training data. My research was focused on leveraging curated monolingual resources for compensating this requirement.
As Einstein said, "Life is like riding a bicycle". So I try to keep moving.
Email:
at
gmail.com