检索结果-内蒙古大学图书馆

8th International Conference on End-User Development, IS-EUD 2021

作者： Fischer, Gerhard Department of Computer Science and Institute of Cognitive Science University of Colorado Boulder United States

ISBN: (纸本)9783030798390

End-User Development (EUD) represents the objective to empower all stakeholders (designers, users, workers, learners, teachers) to actively participate and to make their voices heard in personally meaningful problems. Artificial Intelligence (AI) is currently being considered world-wide as a "deus ex machina"—despite lacking a generally accepted definition, it is credited with miraculous abilities to solve all problems. The presentation will explore and differentiate how AI approaches can support or inhibit different stakeholders to cope with wicked problems in a changing world for which EUD is essential. The relationship between specific AI approaches and meta-design and cultures of participation (being promising frameworks to support EUD) will be explored and critically assessed and prototypical system developments will be described to illustrate different design strategies that will advance EUD not only as a technology, but as a cultural transformation. © 2021, Springer Nature Switzerland AG.

关键词： Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Hybrid approach sentiment analysis using Transformer-LSTM in the Indonesian language

Hybrid approach sentiment analysis using Transformer-LSTM in...

引用

International Seminar on Research of Information Technology and Intelligent Systems (ISRITI)

作者： Angel Karentia Ferdinand Winaya Derwin Suhartono Computer Science Department BINUS Graduate Program – Master of Bina Nusantara University Jakarta Indonesia Computer Science Department School of Computer Science Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798331519643

ISBN: (纸本)9798331519650

Sentiment analysis is widely used as a tool to find valuable insight from texts without explicitly expressed. Lots of techniques have already been used to get it but there still have shortcomings in data source or the model strategy itself. Indonesian language approximately has 199 million speakers across the world where 44 million speakers natively. Even with that great number, the resources for the Indonesian language's natural language processing are still limited, and hard to find the perfect way to define sentiment analysis in the Indonesian language. The state-of-the-art sentiment analysis method uses LSTM on a small corpus while the best in town is Transformer where it's easy to transfer learning from the Transformer pre-trained model into specific tasks. From this potential, combining the Transformer pre-trained model with LSTM can be an innovative strategy. This research compared the hybrid model of Transformer-LSTM build using three Indonesian languages’ Transformer-based pre-trained model on sentiment analysis task which can surpass the Transformer model with the highest increased 2.40% accuracy and 3.02% on F1 score.

关键词： Seminars Sentiment analysis Analytical models Accuracy Law Soft sensors Transfer learning Medical services Transformers Long short term memory

来源：评论

学校读者我要写书评

暂无评论

Neural Machine Translation Approach for Low-resource Languages using Long Short-term Memory Model

Neural Machine Translation Approach for Low-resource Languag...

引用

2023 International Conference on computer science, Information Technology and Engineering, ICCoSITE 2023

作者： Heryadi, Yaya Wijanarko, Bambang Dwi Murad, Dina Fitria Tho, Cuk Hashimoto, Kiyota Bina Nusantara University BINUS Graduate Program - Computer Science Computer Science Department Jakarta11480 Indonesia BINUS Online Learning Computer Science Department Bina Nusantara University Jakarta11480 Indonesia BINUS Online Learning Information Systems Department Bina Nusantara University Jakarta11480 Indonesia Prince of Songkla University Interdisciplinary Graduate School or Earth System Science and Andaman Natural Disaster Management Phuket83120 Thailand

ISBN: (纸本)9798350320954

Communication is a form of translation that human learns naturally since early childhood. Translating a language to another language has become instrumental when peoples interact with other people who speak a different language. One example of high-impact language translation is translating health protocols from the World Health Organization to many national languages to prevent the spread of the Covid-19 virus. However, the language translation is not an easy computation task when there is a language-resource gap. This paper presents empirical results on the performance of the Long Short-term Memory model as a machine translation involving Bahasa Indonesia and the Sundanese languages. Performance of the Sundanese language to Indonesian language translation is 0.93 average training and validation accuracy and 0.91 average testings BLEU score. Whilst, Indonesian language to Sundanese language translation is 0.92 average training and validation accuracy and 0.88 average testings BLEU score. Both models achieve 0.39 average training loss and 0.38 average validation loss. © 2023 IEEE.

关键词： Viruses

来源：评论

学校读者我要写书评

暂无评论

Improved IP lookup technology for trie-based data structures

引用

Journal of computer and System sciences 2023年 133卷 41-55页

作者： Lin, Yen-Heng Hsieh, Sun-Yuan Master Degree Program on Artificial Intelligence National Cheng Kung University Tainan701 Taiwan Department of Computer Science and Information Engineering Institute of Manufacturing Information Systems Institute of Medical Informatics National Cheng Kung University Tainan701 Taiwan Department of Computer Science and Information Engineering National Chi Nan University Nantou Puli Township54561 Taiwan

Many Internet protocol (IP) lookup algorithms have been formulated to improve network performance. This study reviewed and experimentally evaluated technologies for trie-based methods that reduce memory access, memory consumption and IP lookup time. Experiments involving binary tries (for simplicity) were conducted on four real-world data sets, two of which were on IPv4 router tables (855,997 and 876,489 active prefixes) and two of which were on IPv6 router tables (155,310 and 157,579 active prefixes). Technologies designed to reduce time taken (at the expense of memory consumption) worked well for IPv4, and those designed to reduce memory consumption worked well for IPv6. Various combinations of these technologies were applied together in the experiments. © 2022 Elsevier Inc.

关键词： Table lookup

来源：评论

学校读者我要写书评

暂无评论

Robust and Explainable Depression Identification from Speech Using Vowel-Based Ensemble Learning Approaches

arXiv

引用

arXiv 2024年

作者： Feng, Kexin Chaspari, Theodora Department of Computer Science and Engineering Texas A&M University College StationTX United States Institute of Cognitive Science Department of Computer Science University of Colorado Boulder BoulderCO United States

This study investigates explainable machine learning algorithms for identifying depression from speech. Grounded in evidence from speech production that depression affects motor control and vowel generation, pre-trained vowel-based embeddings, that integrate semantically meaningful linguistic units, are used. Following that, an ensemble learning approach decomposes the problem into constituent parts characterized by specific depression symptoms and severity levels. Two methods are explored: a "bottom-up" approach with 8 models predicting individual Patient Health Questionnaire-8 (PHQ-8) item scores, and a "top-down" approach using a Mixture of Experts (MoE) with a router module for assessing depression severity. Both methods depict performance comparable to state-of-the-art baselines, demonstrating robustness and reduced susceptibility to dataset mean/median values. System explainability benefits are discussed highlighting their potential to assist clinicians in depression diagnosis and screening. © 2024, CC BY.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Recurrent Neural Network-based Models as Bahasa Indonesia-Sundanese Language Neural Machine Translator

Recurrent Neural Network-based Models as Bahasa Indonesia-Su...

引用

2023 International Conference on computer science, Information Technology and Engineering, ICCoSITE 2023

作者： Wijanarko, Bambang Dwi Heryadi, Yaya Murad, Dina Fitria Tho, Cuk Hashimoto, Kiyota BINUS Online Learning Computer Science Department Bina Nusantara University Jakarta11480 Indonesia BINUS Graduate Program - Doctor of Computer Science Computer Science Department Bina Nusantara University Jakarta11480 Indonesia BINUS Online Learning Information Systems Department Bina Nusantara University Jakarta11480 Indonesia Prince of Songkla University Interdisciplinary Graduate School or Earth System Science and Andaman Natural Disaster Management Phuket83120 Thailand

ISBN: (纸本)9798350320954

Sequential data are available in many domains including weather observation data, stock price, gene sequence, and many others. Sequential data processing is very instrumental as a solution to a variety of downstream tasks in Natural Language Processing such as language translation. Translating a language to another language has become instrumental when peoples interact with other people who speak a different language. However, language translation is not an easy computation task when there is a language-resource gap. This paper presents some empirical results from the exploration of the Recurrent Neural Network (RNN), the Gated Recurrent Unit (GRU), the Long Short-term Memory (LSTM), and the Bidirectional Long Short-term Memory (Bi-LSTM) models as machine translation models from Bahasa Indonesia to the Sundanese language. These empirical results show that the Bi-LSTM achieved the highest performance (0.95 average training accuracy and 0.92 average testings BLEU score) compared to the RNN (0.93 average training accuracy and 0.91 average testings BLEU score), the GRU (0.88 average training accuracy and 0.89 average testings BLEU score), and the LSTM (0.92 average training accuracy and 0.91 average testings BLEU score) models. These results validate some previously reported studies that claim the Bi-LSTM model potentially outperforms the other recurrent neural network-based models when it is used to process sequence datasets. © 2023 IEEE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Machine Learning-Based Classification of Suicidal Ideation in Tweets: A Sentiment Analysis Approach

Machine Learning-Based Classification of Suicidal Ideation i...

引用

International Conference on Informatics, Multimedia, Cyber and Information System (ICIMCIS)

作者： Orsino Jerrell Bruinier Irma Kartika Wairooy Vladimir Elwin Leonard Brilly Andro Makalew Computer Science Department School of Computer Science Bina Nusantara University Jakarta Indonesia Computer Science Department Mobile Application & Technology Program Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798331533243

ISBN: (纸本)9798331533250

Suicide appears to have reached the highest public health importance, with social media like Twitter providing an infinite source of information that may help in the detection of ideation. This paper describes the development of an effective machine learning-based model for the classification of suicidal ideation in tweets using sentiment analysis. An assembly was done with a dataset of 9,120 tweets; they were labelled according to the suicidal thoughts expressed in them. We presented the assembly and preprocessing of a dataset consisting of 9,120 tweets designated based on indicative or non-indicative suicidal thoughts. The preprocessing steps consisted of normalization, tokenization, stop word removal, and lemmatization. Five ML models, namely Support Vector Machine (SVM), Multinomial Naive Bayes (MNB), Random Forest (RF), Stochastic Gradient Descent (SGD), and Logistic Regression (LR), were further trained on the selected dataset. The models are tested for standard classification metrics: accuracy, precision, recall, F1-Score, and AUC. The best performance for the model was Random Forest with 95.135% and an F1-Score of 95.094%. In conclusion, our study established that an amalgamation of sentiment analysis in ML models can have the potential to detect early suicidal tendencies and that it lays a platform for timely interventions. Future research should be focused on making the dataset even larger and enhancing the detection of sarcasm and jokes in the models.

关键词： Support vector machines Radio frequency Sentiment analysis Analytical models Accuracy Social networking (online) Blogs Training data Random forests Assembly

来源：评论

学校读者我要写书评

暂无评论

Machine Learning-Powered Design and Implementation for Classification of Missing Data in IoT Applications 1

引用

3rd International Conference on Mechanical and Energy Technologies, ICMET 2023

作者： Deepan, S. Suchitra, B. Arya, Namita Ghori, Md. Atheeq Sultan Bora, Ashim Mahender, Nagurla School of CS & AI SR University Telangana Warangal India IT & Cognitive System Tamilnadu Coimbatore India Galgotias University UP Greater Noida India Department of Computer Science and Engineering Telangana University Nizamabad Telangana State Nizamabad India Kampur College Assam Kampur India School of Computer Science and Artificial Intelligence SR University Telangana Warangal India

ISBN: (数字)9789819727162

ISBN: (纸本)9789819727155

Missing values are a common issue in machine learning, which has formed the foundation for data analysis and extraction. Several reasons can lead to missing values, such as missing entirely, missing at random, and not missing. However, before conducting any data analysis, it is imperative to tackle missing values since this might result in biased or inaccurate results. One common technique to increase the precision of IoT data classification is to combine many sensors with different data techniques. But all modalities are seldom available even for the evaluation results, and this lack of data creates serious obstacles to multimodal education. Inspired by recent developments in deep learning, we present a cross-neural network for the segmentation of the IoT Data Classification, which is created on data modalities not all obtainable during trials. We educate the framework in IoT Data Classification using a cost function specifically designed for imbalanced classes. We are giving the device an insufficiently full baseline dataset. Our technique extends beyond CNN training and the gathering of two CNNs trained in the missing modality by using temporal data, considering that they are not involved in the research process. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Cost functions

来源：评论

学校读者我要写书评

暂无评论

Perfomance Improvement of mpdCNN as Ensemble Learning Model on Low Resolution Image-Face Recognition System 8

Perfomance Improvement of mpdCNN as Ensemble Learning Model ...

引用

8th International Conference on Electronics, Materials Engineering and Nano-Technology, IEMENTech 2025

作者： Wangean, Daniel Anando Suhartono, Derwin Bina Nusantara University Master of Computer Science Binus Graduate Program Jakarta Indonesia School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798331520762

Low resolution image-face recognition system is one of the challenging aspects of face recognition models' development. From machine learning, deep learning, and into ensemble learning are implemented to develop face recognition models. Including mpdCNN that published by Mishra in a journal article. This mpdCNN already achieved 88.6% accuracy score on SCFace dataset while published. Based on literature review, ensemble learning model can be improved by modifying parallel layers. The purpose of this research is to improve mpdCNN's performance by implementing some modifications include adding parallel layers, alternating the fusion layers, expanding the layers, and adding residual connections. Some modifications were successfully improved mpdCNN's accuracy score. By adding parallel layers and residual connections into mpdCNN's architecture, the modified mpdCNN that proposed in this research achieved 92.33% accuracy score, measured by Rank-k evaluation metric. © 2025 IEEE.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Revalidating the Encoder-Decoder Depths and Activation Function to Find Optimum Vanilla Transformer Model

Revalidating the Encoder-Decoder Depths and Activation Funct...

引用

2023 International Conference on computer science, Information Technology and Engineering, ICCoSITE 2023

作者： Heryadi, Yaya Wijanarko, Bambang Dwi Fitria Murad, Dina Tho, Cuk Hashimoto, Kiyota Bina Nusantara University Graduate Program- Computer Science Computer Science Department BINUS Jakarta11480 Indonesia Bina Nusantara University Online Learning Computer Science Department BINUS Jakarta11480 Indonesia Bina Nusantara University BINUS Online Learning Information Systems Department Jakarta11480 Indonesia Prince of Songkla University Interdisciplinary Graduate School or Earth System Science and Andaman Natural Disaster Management Phuket83120 Thailand

ISBN: (纸本)9798350320954

The transformer model has become a state-of-the-art model in Natural Language Processing. The initial transformer model, known as the vanilla transformer model, is designed to improve some prominent models in sequence modeling and transduction problems such as language modeling and machine translation. The initial transformer model has 6 stacks of identical encoder-decoder layers with an attention mechanism whose aim is to push limitations of common recurrent language models and encoder-decoder architectures. Its outstanding performance has inspired many researchers to extend the architecture to improve its performance and computation efficiency. Despite many extensions to the vanilla transformer, there is no clear explanation of the encoder-decoder set out depth in the vanilla transformer model. This paper presents exploration results on the effect of combination encoder-decoder layer depth and activation function in the feed-forward layer of the vanilla transformer model on its performance. The model is tested to address a downstream task: text translation from Bahasa Indonesia to the Sundanese language. Although the value difference is not significantly large, the empirical results show that the combination of depth = 2 with Sigmoid, Tanh, and ReLU activation function;and d = 6 with ReLU activation shows the highest average training accuracy. Interestingly, d = 6 and ReLU show the lowest average training and validation loss. However, statistically, there is no significant difference between depth and activation functions. © 2023 IEEE.

关键词： Neural machine translation

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：