检索结果-内蒙古大学图书馆

Count-Data Mixed Models of Topical Tweets: A Case of Indonesia Flood Events

Procedia computer science 2023年 227卷 805-812页

作者： Alam Ahmad Hidayat Bens Pardamean Mathematics Department School of Computer Science Bina Nusantara University Jakarta Indonesia 11480 Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia 11480 Computer Science Department BINUS Graduate Program – Master of Computer Science Program Bina Nusantara University Jakarta Indonesia 11480

Examining topic-level variability in modeling Twitter data can potentially yield more comprehensive insights into public perception during critical periods, thereby enhancing natural disaster mitigation and surveillance efforts. In this study, we utilized generalized linear mixed models (GLMMs) to illustrate the variability in tweet counts related to specific topics in Indonesia during the flood events that occurred in February 2021. The glmmTMB library in R was employed for this purpose. The data were assumed to follow two distinct exponential distributions: Poisson and Negative Binomial. To incorporate random effects, random intercepts and random slopes were introduced, allowing them to vary randomly across topics in the initial two models. Additionally, the final model addressed issues related to dispersion and zero-inflation. By evaluating the Akaike Information Criteria scores, we determined that a model based on the Negative Binomial distribution with random zero-inflation intercepts best fit the data. The chosen model formulation and the estimated parameters have the potential to forecast topic-specific trends in Indonesian flood-related Twitter data.

关键词： count data disaster surveillance floods mixed models Twitter

来源：评论

学校读者我要写书评

暂无评论

Indonesian Agricultural-crops Classification Using Transfer Learning Model

引用

Procedia computer science 2023年 227卷 128-136页

作者： Mahmud Isnan Alam Ahmad Hidayat Bens Pardamean Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta 11480 Indonesia Department of Mathematics School of Computer Science Bina Nusantara University Jakarta 11480 Indonesia Computer Science Department BINUS Graduate Program – Master of Computer Science Program Bina Nusantara University Jakarta 11480 Indonesia

Classification of Indonesian crops is a critical task in developing farming and getting more understanding of agriculture. However, there is no clear task in classifying types of crops in Indonesia. Transfer learning has been used successfully in a variety of image classification applications. Thus, in this paper, we collected images of Indonesian crops from the internet randomly and proposed a classification by using transfer learning of deep learning with four pre-trained models: EffficientNet- B0, ResNet18, VGG19, and AlexNet. In the experiment, augmentation techniques such as random horizontal flip, random vertical flip, and random affine were utilized to prevent the network from overfitting. The result found that EfficientNet-B0 outperformed other models with an accuracy of 82.55. Then, the model struggled to distinguish between crops in the same family. According to the results, although transfer learning can work well to classify images of Indonesian agricultural crops, some improvements are still required to address existing issues.

关键词： Agriculture Classification Crop Deep Learning Transfer learning

来源：评论

学校读者我要写书评

暂无评论

A Hybrid CNN-SVM Approach for Malaria in Blood Cell Detection

A Hybrid CNN-SVM Approach for Malaria in Blood Cell Detectio...

引用

Software Engineering and Information Technology (ICoSEIT), International Conference on

作者： Andrea Stevens Karnyoto Mahmud Isnan Bens Pardamean Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia Computer Science Department BINUS Graduate Program - Master of Computer Science Bina Nusantara University Jakarta Indonesia

There is still a severe malaria problem worldwide, particularly in regions with limited access to diagnostic tools. It is crucial to develop a system for detecting malaria in blood cells. This paper presents a hybrid Convolutional Neural Network (CNN), and as a classifier, we use a Support Vector Machine (SVM) framework for the automated detection of malaria parasites in blood cell images. The proposed system leverages the strengths of CNNs in feature extraction and representation learning from images, combined with the discriminative power of SVMs for classification. Initially, CNN extracts intricate features from blood cell images, capturing essential patterns indicative of malaria infection. Subsequently, the extracted features are used to train an SVM classifier, enabling accurate discrimination between parasitized and uninfected blood cells. Experimental dataset evaluations were obtained from the Lister Hill National Center for Biomedical Communications website of the National Library of Medicine. The proposed model achieves a better f1-score, outperforming individual CNN or SVM models, around 0.015 compared to individual CNN models and 0.27 compared to individual SVM models. This hybrid CNN-SVM methodology offers a promising solution for accurately and efficiently detecting malaria parasites in blood cell images.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Leveraging ChemBERTa Robustness in Drug-Drug Interaction Classification via Molecular Decomposition

Leveraging ChemBERTa Robustness in Drug-Drug Interaction Cla...

引用

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

作者： Christian Kenneth Alyssa Imani Bens Pardamean Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia Computer Science Department BINUS Graduate Program – Master of Computer Science Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798331519643

ISBN: (纸本)9798331519650

Understanding drug-drug interactions (DDIs) can benefit us by enhancing drug effectiveness and preventing side effects that arise from consuming multiple drugs administered together. However, as the number of novel drugs increases, predicting potential DDIs becomes more challenging. Previous studies have used the molecular language model ChemBERTa to predict potential DDIs, but the predictions were only tested on unseen drugs and not dissimilar drugs, while also performing poorly. In this study, we demonstrate that a simple molecular preprocessing method using BRICS decomposition can enhance ChemBERTa robustness in predicting potential DDIs for unseen-dissimilar drugs. Specifically, we employ fragmentation with a minimum fragment size of 5 (BRICS-5), which balances the number of fragments and decomposition speed. Using this method, we were able to improve ChemBERTa robustness, reflected by an approximate increase of 1.5% in recall and 2.2% in the f1-score compared to non-decomposed molecules. Our findings suggest that this preprocessing method could be useful in the development of future AI-based approaches for solving challenging molecular tasks across different settings and various model architectures.

关键词： Drugs Seminars Accuracy Data preprocessing Predictive models Robustness Data models Intelligent systems Information technology

来源：评论

学校读者我要写书评

暂无评论

Implementation of Fine-Tuned BERT for Enzyme Classification Based on Gene Ontology

Implementation of Fine-Tuned BERT for Enzyme Classification ...

引用

International Conference of Advanced Informatics: Concept, Theory and Application (ICAICTA)

作者： Matthew Martianus Henry Christian Kenneth Bens Pardamean Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia Computer Science Department BINUS Graduate Program - Master of Computer Science Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798331520311

ISBN: (纸本)9798331520328

Enzymes are biocatalysts with vital roles in biological functions and many industrial applications. Diverse enzymes are classified using Enzyme Commission (EC) nomenclature, making differentiation challenging. On the other hand, another biological information, gene ontology (GO), can describe the biological aspects of enzymes, covering related biological processes (BP), molecular functions (MF), and their locations within cells (CC). This study proposes a novel EC class and subclass classification of enzymes within the ontology subclass based on their GO semantics using a Bidirectional Encoder Representation of Transformer (BERT). The BERT model is first fine-tuned using the preprocessed GO term name and definition, with the enzymes in each ontology class (BP, MF, or CC) are also divided based on how the GO assigned, either through manual annotation (NONIEA) or electronically inferred (IEA). BERT successfully obtained 0.93, 0.60, 0.99, 0.90, 0.40, and 0.35 F1 scores during fine-tuning for BP IEA, BP NONIEA, MF IEA, MF NONIEA, CC IEA, and CC NONIEA, respectively. On the test set, the fine-tuned BERT significantly outperformed GOntoSim, a framework to calculate semantic similarity based on classical information theory, in EC class classification across all metrics with less inference time in all ontology subclass. Expanded further to the EC subclass, BERT can classify the enzyme on the EC subclass level in BP IEA and MF IEA ontology subclass. However, longer epochs are needed in fine-tuning. This result shows that the names and definitions of GO terms are distinguishable features in classifying enzymes as an alternative to the information content approach.

关键词： Measurement Enzymes Semantics Bidirectional control Manuals Ontologies Transformers Encoding Biological information theory Informatics

来源：评论

学校读者我要写书评

暂无评论

Explainable Tree-Based Boosting Algorithms for Smoker Detection Using Bio-Signals

Explainable Tree-Based Boosting Algorithms for Smoker Detect...

引用

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

作者： Jason Sebastian Sulistyawan Kuncahyo Setyo Nugroho Bens Pardamean Computer Science Department BINUS Graduate Program Master of Computer Science Bina Nusantara University Jakarta Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798331519643

ISBN: (纸本)9798331519650

Many smoking-related diseases are difficult to treat and often fatal. Rather than treating diseased smokers, preventing the diseased is more achievable, though, many of them deny to being smokers, leading to another problem. Thus, this study aims to detect important aspects that can detect that the person is a smoker or not through their bio-signals through using SHAP, along with a comprehensive analysis of the used methods, gradient-boosting algorithms XGBoost, LightGBM, and CatBoost, known for their efficiency in handling complex datasets and non-linear relationships. The study then found that triglyceride, Gtp, hemoglobins significantly affect the body's responses to smoking, based on the CatBoosts’ results, having an AUC score of up to 0.8612 and an accuracy score of up to 0.7776 with the selected features.

关键词： Seminars Measurement Accuracy Predictive models Feature extraction Boosting Physiology Classification algorithms Reliability Diseases

来源：评论

学校读者我要写书评

暂无评论

Systematic Literature Review on Metaheuristic Algorithms in SNP Analysis for Diseases

Systematic Literature Review on Metaheuristic Algorithms in ...

引用

International Conference on Cybernetics and Intelligent System (ICORIS)

作者： Jason Sebastian Sulistyawan Kuncahyo Setyo Nugroho Bens Pardamean Computer Science Department BINUS Graduate Program - Master of Computer Science Bina Nusantara University Jakarta Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798331539603

ISBN: (纸本)9798331539610

Single-nucleotide polymorphism (SNP) analysis has become a pivotal strategy for drug discovery within bioinformatics, especially for incurable diseases like cancer. With the increasing number of researchers starting to embrace metaheuristic methods, a review was done by gathering papers from the Google Scholar database from 2018 to 2023, which resulted in 20 papers after title, abstract, and content filtering. The findings show that the Genetic Algorithm and the Harmony Search Algorithm have become popular approaches in SNP analysis, particularly in studies on breast cancer, age-related macular degeneration, and colorectal cancer. However, the review shows that while researchers have proven most methods effective in finding disease-related SNPs, a more measurable study in SNP analysis is needed, due to the lack of elaboration on measurement metrics in the found studies.

关键词： Macular degeneration Filtering Databases Metaheuristics Time measurement Internet Drug discovery Intelligent systems Genetic algorithms Systematic literature review

来源：评论

学校读者我要写书评

暂无评论

Knowledge Graph-Enhanced Semantic Cache for Low-Latency and Cost-Effective Inference in Large Language Models

Knowledge Graph-Enhanced Semantic Cache for Low-Latency and ...

引用

International Conference on Information Management and Technology (ICIMTech)

作者： Nicholas Dominic Bens Pardamean Bioinformatics & Data Science Research Center Bina Nusantara University Jakarta Indonesia Computer Science Department BINUS Graduate Program - Master of Computer Science Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798350390025

ISBN: (纸本)9798350390032

In organizational knowledge management, Large Language Model (LLM) caches act as a semantic repository gathered from previous LLM responses. Due to intensive calls from multiple users, LLM may suffer from high inference latency. While there are many prior available approaches to solve this problem, most of them are inherently complex. This paper introduced a Knowledge Graph-enhanced Semantic Cache mechanism as an alternative, lightweight technique to boost retrieval for similar prompts. The latest state-of-the-art open-source LLM, named Google's Gemma-2B-it, was used to generate sample prompts and responses as a draft, while a knowledge graph (KG) was built from Wikipedia sentences. To create embeddings of prompts and KG, all-MiniLM-L6-v2 from SentenceTransformer was used. This new cache system resulted in up to 28% improvement over a standard model. In particular, reinforcement with KG cache embeddings yielded more than 85% semantic cache accuracy. To map the next trajectory of this pilot study, an overview of the extended framework for LLM knowledge management was also presented in this paper. The framework includes the new KG- enhanced cache system equipped with scalable security and fallback mechanisms that can promote green technology through substantial improvements in latency, throughput, and overall LLM costs.

关键词： Large language models Semantics Throughput Knowledge management Internet Trajectory Security Online services Low latency communication Standards

来源：评论

学校读者我要写书评

暂无评论

Long Short-Term Memory-based Models for Sleep Quality Prediction from Wearable Device Time Series Data

引用

Procedia computer science 2023年 227卷 1062-1069页

作者： Alam Ahmad Hidayat Arif Budiarto Bens Pardamean Mathematics Department School of Computer Science Bina Nusantara University Jakarta 11480 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta 11480 Indonesia Computer Science Department School of Computer Science Bina Nusantara University Jakarta 11480 Indonesia Computer Science Department BINUS Graduate Program – Master of Computer Science Program Bina Nusantara University Jakarta 11480 Indonesia

Several studies suggest that sleep quality is associated with physical activities. Moreover, deep sleep time can be used to determine the sleep quality of an individual. In this work, we aim to find the association between physical activities and deep sleep time by modeling the time series data such as heart rate and a number of steps captured from a commercial wearable device. Our previous study demonstrates that deep learning-based time series modeling is well suited for our problem since the temporal patterns in the two physical parameters need to be captured to obtain more accurate results. We first preprocess our series data to have a time-step size of 10 minutes. To improve our previous effort in this modeling, we compare four different variants of Long Short-Term Memory (LSTM)-based models, ranging from single input to dual input models. Our result shows that the simple stacked LSTM model performs better for our data because the remaining models suffer from overfitting due to a larger number of the trained parameters.

关键词： deep learning forecasting sleep quality time series wearable devices

来源：评论

学校读者我要写书评

暂无评论

Mitigating Berkson's Paradox with Neural Propensity Score Matching in E-Commerce Deals

Mitigating Berkson's Paradox with Neural Propensity Score Ma...

引用

International Conference on Cybernetics and Intelligent System (ICORIS)

作者： Nicholas Dominic Bens Pardamean Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia Computer Science Department BINUS Graduate Program – Master of Computer Science Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798331539603

ISBN: (纸本)9798331539610

Within observational Data science workloads, Berkson's paradox can lead to false causal inferences. One of the prominent quasi-experimental methods to mitigate this selection bias is Propensity Score Matching (PSM). An approach called Neural PSM (NPSM) was developed to overcome the drawbacks of conventional regression-based PSM, including its limited flexibility to model high-dimensional data and non-linear relationships that could cause imperfect covariate balance. In this study, a three-layer depth of Deep Neural Networks was designed to estimate propensity scores and finally balance both control and treatment groups of the Groupon dataset. An unsupervised k-Nearest Neighbor algorithm then helped the model to efficiently detect and cluster similar matching points. From the five salient features presented, NPSM successfully achieved lower differences in Cohen's d effect size, i.e., 0.313 for coupon duration, 0.017 for promotion length, 0.425 for quantity sold, -0.199 for limited supply, and 0.395 for Facebook likes. While these results mostly outperformed Linear Regression (LR) and Random Forest (RF) models, further evaluation is needed to verify the true effectiveness of NPSM in mitigating Berkson's paradox in broader e-commerce contexts.

关键词： Radio frequency Social networking (online) High dimensional data Linear regression Nearest neighbor methods Data science Electronic commerce Intelligent systems Cybernetics Context modeling

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：