Search Results for author: Yoshimasa Tsuruoka

Found 50 papers, 12 papers with code

WSPAlign: Word Alignment Pre-training via Large-Scale Weakly Supervised Span Prediction

2 code implementations9 Jun 2023 Qiyu Wu, Masaaki Nagata, Yoshimasa Tsuruoka

Most existing word alignment methods rely on manual alignment datasets or parallel corpora, which limits their usefulness.

Word Alignment

Unsupervised Discovery of Continuous Skills on a Sphere

no code implementations21 May 2023 Takahisa Imagawa, Takuya Hiraoka, Yoshimasa Tsuruoka

However, most of the existing methods learn a finite number of discrete skills, and thus the variety of behaviors that can be exhibited with the learned skills is limited.

Unsupervised Reinforcement Learning

Which Experiences Are Influential for Your Agent? Policy Iteration with Turn-over Dropout

1 code implementation26 Jan 2023 Takuya Hiraoka, Takashi Onishi, Yoshimasa Tsuruoka

In reinforcement learning (RL) with experience replay, experiences stored in a replay buffer influence the RL agent's performance.

reinforcement-learning Reinforcement Learning (RL)

Soft Sensors and Process Control using AI and Dynamic Simulation

no code implementations8 Aug 2022 Shumpei Kubosawa, Takashi Onishi, Yoshimasa Tsuruoka

During the operation of a chemical plant, product quality must be consistently maintained, and the production of off-specification products should be minimized.

EASE: Entity-Aware Contrastive Learning of Sentence Embedding

1 code implementation NAACL 2022 Sosuke Nishikawa, Ryokan Ri, Ikuya Yamada, Yoshimasa Tsuruoka, Isao Echizen

We present EASE, a novel method for learning sentence embeddings via contrastive learning between sentences and their related entities.

Clustering Contrastive Learning +5

Pretraining with Artificial Language: Studying Transferable Knowledge in Language Models

no code implementations ACL 2022 Ryokan Ri, Yoshimasa Tsuruoka

We investigate what kind of structural knowledge learned in neural network encoders is transferable to processing natural language.

Railway Operation Rescheduling System via Dynamic Simulation and Reinforcement Learning

no code implementations17 Jan 2022 Shumpei Kubosawa, Takashi Onishi, Makoto Sakahara, Yoshimasa Tsuruoka

The system leverages reinforcement learning and a dynamic simulator that can simulate the railway traffic and passenger flow of a whole line.

reinforcement-learning Reinforcement Learning (RL) +1

A Multilingual Bag-of-Entities Model for Zero-Shot Cross-Lingual Text Classification

no code implementations15 Oct 2021 Sosuke Nishikawa, Ikuya Yamada, Yoshimasa Tsuruoka, Isao Echizen

We present a multilingual bag-of-entities model that effectively boosts the performance of zero-shot cross-lingual text classification by extending a multilingual pre-trained language model (e. g., M-BERT).

Entity Typing Language Modelling +3

mLUKE: The Power of Entity Representations in Multilingual Pretrained Language Models

2 code implementations ACL 2022 Ryokan Ri, Ikuya Yamada, Yoshimasa Tsuruoka

We train a multilingual language model with 24 languages with entity representations and show the model consistently outperforms word-based pretrained models in various cross-lingual transfer tasks.

 Ranked #1 on Cross-Lingual Question Answering on XQuAD (Average F1 metric, using extra training data)

Cross-Lingual Question Answering Cross-Lingual Transfer +1

Dropout Q-Functions for Doubly Efficient Reinforcement Learning

2 code implementations ICLR 2022 Takuya Hiraoka, Takahisa Imagawa, Taisei Hashimoto, Takashi Onishi, Yoshimasa Tsuruoka

To make REDQ more computationally efficient, we propose a method of improving computational efficiency called DroQ, which is a variant of REDQ that uses a small ensemble of dropout Q-functions.

Q-Learning reinforcement-learning +1

Zero-pronoun Data Augmentation for Japanese-to-English Translation

no code implementations ACL (WAT) 2021 Ryokan Ri, Toshiaki Nakazawa, Yoshimasa Tsuruoka

For Japanese-to-English translation, zero pronouns in Japanese pose a challenge, since the model needs to infer and produce the corresponding pronoun in the target side of the English sentence.

Data Augmentation Machine Translation +1

Modeling Target-side Inflection in Placeholder Translation

1 code implementation MTSummit 2021 Ryokan Ri, Toshiaki Nakazawa, Yoshimasa Tsuruoka

Placeholder translation systems enable the users to specify how a specific phrase is translated in the output sentence.

LEMMA Translation

Utilizing Skipped Frames in Action Repeats via Pseudo-Actions

no code implementations7 May 2021 Taisei Hashimoto, Yoshimasa Tsuruoka

The key idea of our method is making the transition between action-decision points usable as training data by considering pseudo-actions.

Continuous Control OpenAI Gym +2

Meta-Model-Based Meta-Policy Optimization

no code implementations4 Jun 2020 Takuya Hiraoka, Takahisa Imagawa, Voot Tangkaratt, Takayuki Osa, Takashi Onishi, Yoshimasa Tsuruoka

Model-based meta-reinforcement learning (RL) methods have recently been shown to be a promising approach to improving the sample efficiency of RL in multi-task settings.

Continuous Control Meta-Learning +3

Data Augmentation with Unsupervised Machine Translation Improves the Structural Similarity of Cross-lingual Word Embeddings

no code implementations ACL 2021 Sosuke Nishikawa, Ryokan Ri, Yoshimasa Tsuruoka

Unsupervised cross-lingual word embedding (CLWE) methods learn a linear transformation matrix that maps two monolingual embedding spaces that are separately trained with monolingual corpora.

Cross-Lingual Word Embeddings Data Augmentation +3

Revisiting the Context Window for Cross-lingual Word Embeddings

no code implementations ACL 2020 Ryokan Ri, Yoshimasa Tsuruoka

Existing approaches to mapping-based cross-lingual word embeddings are based on the assumption that the source and target embedding spaces are structurally similar.

Bilingual Lexicon Induction Cross-Lingual Word Embeddings +1

Optimistic Proximal Policy Optimization

no code implementations25 Jun 2019 Takahisa Imagawa, Takuya Hiraoka, Yoshimasa Tsuruoka

Reinforcement Learning, a machine learning framework for training an autonomous agent based on rewards, has shown outstanding results in various domains.

BIG-bench Machine Learning reinforcement-learning +1

Building a Computer Mahjong Player via Deep Convolutional Neural Networks

no code implementations5 Jun 2019 Shiqi Gao, Fuminori Okuya, Yoshihiro Kawahara, Yoshimasa Tsuruoka

The evaluation function for imperfect information games is always hard to define but owns a significant impact on the playing strength of a program.

Game of Go Test

Learning Robust Options by Conditional Value at Risk Optimization

1 code implementation NeurIPS 2019 Takuya Hiraoka, Takahisa Imagawa, Tatsuya Mori, Takashi Onishi, Yoshimasa Tsuruoka

While there are several methods to learn options that are robust against the uncertainty of model parameters, these methods only consider either the worst case or the average (ordinary) case for learning options.

Synthesizing Chemical Plant Operation Procedures using Knowledge, Dynamic Simulation and Deep Reinforcement Learning

no code implementations6 Mar 2019 Shumpei Kubosawa, Takashi Onishi, Yoshimasa Tsuruoka

Chemical plants are complex and dynamical systems consisting of many components for manipulation and sensing, whose state transitions depend on various factors such as time, disturbance, and operation procedures.

reinforcement-learning Reinforcement Learning (RL) +1

Neural Fictitious Self-Play on ELF Mini-RTS

no code implementations6 Feb 2019 Keigo Kawamura, Yoshimasa Tsuruoka

Despite the notable successes in video games such as Atari 2600, current AI is yet to defeat human champions in the domain of real-time strategy (RTS) games.

reinforcement-learning Reinforcement Learning (RL)

Partially Non-Recurrent Controllers for Memory-Augmented Neural Networks

no code implementations30 Dec 2018 Naoya Taguchi, Yoshimasa Tsuruoka

Memory-Augmented Neural Networks (MANNs) are a class of neural networks equipped with an external memory, and are reported to be effective for tasks requiring a large long-term memory and its selective use.

Refining Manually-Designed Symbol Grounding and High-Level Planning by Policy Gradients

no code implementations29 Sep 2018 Takuya Hiraoka, Takashi Onishi, Takahisa Imagawa, Yoshimasa Tsuruoka

In this paper, we propose a framework that can automatically refine symbol grounding functions and a high-level planner to reduce human effort for designing these modules.

Decision Making

Multilingual Extractive Reading Comprehension by Runtime Machine Translation

1 code implementation10 Sep 2018 Akari Asai, Akiko Eriguchi, Kazuma Hashimoto, Yoshimasa Tsuruoka

Given a target language without RC training data and a pivot language with RC training data (e. g. English), our method leverages existing RC resources in the pivot language by combining a competitive RC model in the pivot language with an attentive Neural Machine Translation (NMT) model.

Machine Translation NMT +2

Monte Carlo Tree Search with Scalable Simulation Periods for Continuously Running Tasks

no code implementations7 Sep 2018 Seydou Ba, Takuya Hiraoka, Takashi Onishi, Toru Nakata, Yoshimasa Tsuruoka

The evaluation results show that, with variable simulation times, the proposed approach outperforms the conventional MCTS in the evaluated continuous decision space tasks and improves the performance of MCTS in most of the ALE tasks.

Atari Games

Accelerated Reinforcement Learning for Sentence Generation by Vocabulary Prediction

1 code implementation NAACL 2019 Kazuma Hashimoto, Yoshimasa Tsuruoka

A major obstacle in reinforcement learning-based sentence generation is the large action space whose size is equal to the vocabulary size of the target-side language.

Image Captioning Machine Translation +4

Hierarchical Reinforcement Learning with Abductive Planning

no code implementations28 Jun 2018 Kazeto Yamamoto, Takashi Onishi, Yoshimasa Tsuruoka

One potential solution to this problem is to combine reinforcement learning with automated symbol planning and utilize prior knowledge on the domain.

Hierarchical Reinforcement Learning reinforcement-learning +1

Neural Machine Translation with Source-Side Latent Graph Parsing

no code implementations EMNLP 2017 Kazuma Hashimoto, Yoshimasa Tsuruoka

This paper presents a novel neural machine translation model which jointly learns translation and source-side latent graph representations of sentences.

Machine Translation NMT +1

Character-based Decoding in Tree-to-Sequence Attention-based Neural Machine Translation

no code implementations WS 2016 Akiko Eriguchi, Kazuma Hashimoto, Yoshimasa Tsuruoka

This paper reports our systems (UT-AKY) submitted in the 3rd Workshop of Asian Translation 2016 (WAT{'}16) and their results in the English-to-Japanese translation task.

Machine Translation NMT +1

Domain Adaptation for Neural Networks by Parameter Augmentation

no code implementations WS 2016 Yusuke Watanabe, Kazuma Hashimoto, Yoshimasa Tsuruoka

Recently, recurrent neural networks have been shown to be successful on a variety of NLP tasks such as caption generation; however, the existing domain adaptation techniques are limited to (1) tune the model parameters by the target dataset after the training by the source dataset, or (2) design the network to have dual output, one for the source domain and the other for the target domain.

Domain Adaptation

Asymmetric Move Selection Strategies in Monte-Carlo Tree Search: Minimizing the Simple Regret at Max Nodes

no code implementations8 May 2016 Yun-Ching Liu, Yoshimasa Tsuruoka

We develop the Asymmetric-MCTS algorithm, which is an MCTS variant that applies a simple regret algorithm on max nodes, and the UCB algorithm on min nodes.

Adaptive Joint Learning of Compositional and Non-Compositional Phrase Embeddings

no code implementations ACL 2016 Kazuma Hashimoto, Yoshimasa Tsuruoka

We present a novel method for jointly learning compositional and non-compositional phrase embeddings by adaptively weighting both types of embeddings using a compositionality scoring function.

Tree-to-Sequence Attentional Neural Machine Translation

1 code implementation ACL 2016 Akiko Eriguchi, Kazuma Hashimoto, Yoshimasa Tsuruoka

Most of the existing Neural Machine Translation (NMT) models focus on the conversion of sequential data and do not directly use syntactic information.

Machine Translation NMT +1

Adapting Improved Upper Confidence Bounds for Monte-Carlo Tree Search

no code implementations11 May 2015 Yun-Ching Liu, Yoshimasa Tsuruoka

The UCT algorithm, which combines the UCB algorithm and Monte-Carlo Tree Search (MCTS), is currently the most widely used variant of MCTS.

Cannot find the paper you are looking for? You can Submit a new open access paper.