检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Schreyer, W. Max Anderson, Christopher Thompson, Reid F. Computational Biology Program Oregon Health & Science University PortlandOR United States Department of Biomedical Engineering Oregon Health & Science University PortlandOR United States Portland VA Research Foundation PortlandOR United States Department of Radiation Medicine Oregon Health & Science University PortlandOR United States Department of Medical Informatics and Clinical Epidemiology Oregon Health & Science University PortlandOR United States Division of Hospital and Specialty Medicine VA Portland Healthcare System PortlandOR United States

Failure of machine learning models to generalize to new data is a core problem limiting the reliability of AI systems, partly due to the lack of simple and robust methods for comparing new data to the original training dataset. We propose a standardized approach for assessing data similarity in a model-agnostic manner by constructing a supervised autoencoder for generalizability estimation (SAGE). We compare points in a low-dimensional embedded latent space, defining empirical probability measures for k-Nearest Neighbors (kNN) distance, reconstruction of inputs and task-based performance. As proof of concept for classification tasks, we use MNIST and CIFAR-10 to demonstrate how an ensemble output probability score can separate deformed images from a mixture of typical test examples, and how this SAGE score is robust to transformations of increasing severity. As further proof of concept, we extend this approach to a regression task using non-imaging data (UCI Abalone). In all cases, we show that out-of-the-box model performance increases after SAGE score filtering, even when applied to data from the model’s own training and test datasets. Our out-of-distribution scoring method can be introduced during several steps of model construction and assessment, leading to future improvements in responsible deep learning implementation. Copyright © 2025, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Integration of HTML tables in web pages

Integration of HTML tables in web pages

引用

International Conference on Data and Software engineering, ICODSE 2015

作者： Akbar, Memen Azizah, Fazat Nur Putri Saptawati, G.A. Magister in Informatics Study Program School of Electrical Engineering and Informatics Institut Teknologi Bandung Bandung Indonesia Data and Software Engineering Research Group School of Electrical Engineering and Informatics Institut Teknologi Bandung Bandung Indonesia

ISBN: (纸本)9781467384285

The growing number of web pages on the internet introduces a need to combine and integrate information from HTML tables of different web pages that contain similar information into a single web page, especially information from the same domain of interest. This paper presents an approach of HTML table integration by combining several existing methods that are proved to solve different issues in the integration processes. The integration of HTML table consists of three phases: (1) extraction of the structure of the tables;(2) integration of the tables' schema;(3) integration of the data values. To solve the conflicts in semantics and naming in the tables schema, domain-ontology is used. To improve quality of integration of data values in the tables, the vector space model is used to check the duplications of data values. As the integration result, a single HTML table is obtained. The approach is implemented on an engine built using Phyton. Results of the experiment shows that the engine can successfully integrate two HTML tables into single table. © 2015 IEEE.

关键词： Websites

来源：评论

学校读者我要写书评

暂无评论

Development Of Scholarship Sustainability Analysis Application Every Semester On STMIK Pelita Nusantara Students 4

Development Of Scholarship Sustainability Analysis Applicati...

引用

4th IEEE International Conference of Computer Science and Information Technology, ICOSNIKOM 2022

作者： Br Tarigan, Nera Mayana Sinaga, Bosker Panggabean, Erwin Sagala, Jijon Raphita Informatics Engineering Study Program Stmik Pelita Nusantara Medan Indonesia Software Engineering Study Program Stmik Pelita Nusantara Medan Indonesia

ISBN: (纸本)9798350399073

Decision making in an organization is needed a system. Decision support system is a system that helps company management and education in decision making that provides information, modeling and analyzing data with the application of a method. The management, especially higher education, must maximize sophisticated technology. For example, to analyze scholarship data. Problems are found in manual data analysis so that human errors often occur in addition to using more time than the target, losing old data, resulting in inaccurate analysis of scholarship sustainability every semester. The research was conducted to overcome the problems that have been found by analyzing the existing data with the simple additive weighting method, building a system by applying the method and testing the system. The result of this research is to produce an application for the development of a decision support system in managing student data to determine the sustainability of the scholarship every semester. Applications that are built can help the management in managing scholarships. © 2022 IEEE.

关键词： Decision support systems

来源：评论

学校读者我要写书评

暂无评论

Development of attitude assessment instrument in engineering mathematics 1 course to assess discussion on MOOC platform 7

Development of attitude assessment instrument in engineering...

引用

7th International Conference on Mathematics, Science, and Education 2020, ICMSE 2020

作者： Yuniwati, I. Yustita, A.D. Hardiyanti, S.A. Suardinata, I.W. Mechanical Engineering Program Mechanical Engineering Department Politeknik Negeri Banyuwangi East Java Banyuwangi Indonesia Management Business of Tourism Program Informatics Engineering Department Politeknik Negeri Banyuwangi East Java Banyuwangi Indonesia Civil Engineering Program Civil Engineering Department Politeknik Negeri Banyuwangi Kabat East Java Banyuwangi Indonesia Informatics Enggineering Program Informatics Engineering Department Politeknik Negeri Banyuwangi East Java Banyuwangi Indonesia

Affective assessment instruments have been developed, especially those related to attitudes. In engineering mathematics learning 1 through the MOOC platform, it was carried out with discussion. The development of attitude instrument assessment to assess student activity was carried out after learning was complete. This development research used the Plomp Model. The Plomp Model consisted of 5 stages. They were the initial investigative stages, design stage;construction stage;the test, evaluation, revision stages;and implementation stages. Two experts and trials carried out validation were carried out on 100 students. The result of developing an attitude assessment instrument was a questionnaire. The difference from other questionnaires lies in paying attention to the internal continuum interval of attitudes developed by Bloom's Taxonomy. They were Responding, Valuing, and Organization. Afterward, The instrument is tested for reliability and validity until it is declared to validate each item and the instrument's reliability in the excellent category. The value of Cronbach's Alpha instrument is 0.848. The value indicates that the attitude assessment instrument is classified as having good criteria. Then the instrument is implemented in discussion on the MOOC Platform. © Published under licence by IOP Publishing Ltd.

关键词： Surveys

来源：评论

学校读者我要写书评

暂无评论

Understanding and controlling the geometry of memory organization in RNNs

arXiv

引用

arXiv 2025年

作者： Haputhanthri, Udith Storan, Liam Jiang, Yiqi Raheja, Tarun Shai, Adam Akengin, Orhun Miolane, Nina Schnitzer, Mark J. Dinc, Fatih Tanaka, Hidenori CNC Program Stanford University United States James H. Clark Center for Biomedical Engineering & Sciences Stanford University United States Kipo AI San Francisco United States Simplex Astera Institute EmeryvilleCA United States Geometric Intelligence Lab UC Santa Barbara Santa Barbara United States Howard Hughes Medical Institute United States Physics & Informatics Laboratories NTT Research Inc. SunnyvaleCA94085 United States Kavli Institute for Theoretical Physics UC Santa Barbara United States CBS-NTT Program in Physics of Intelligence Harvard University CambridgeMA94305 United States

Training recurrent neural networks (RNNs) is a high-dimensional process that requires updating numerous parameters. Therefore, it is often difficult to pinpoint the underlying learning mechanisms. To address this challenge, we propose to gain mechanistic insights into the phenomenon of abrupt learning by studying RNNs trained to perform diverse short-term memory tasks. In these tasks, RNN training begins with an initial search phase. Following a long period of plateau in accuracy, the values of the loss function suddenly drop, indicating abrupt learning. Analyzing the neural computation performed by these RNNs reveals geometric restructuring (GR) in their phase spaces prior to the drop. To promote these GR events, we introduce a temporal consistency regularization that accelerates (bioplausible) training, facilitates attractor formation, and enables efficient learning in strongly connected networks. Our findings offer testable predictions for neuroscientists and emphasize the need for goal-agnostic secondary mechanisms to facilitate learning in biological and artificial networks. © 2025, CC BY.

关键词： Phase space methods

来源：评论

学校读者我要写书评

暂无评论

Double exponential smoothing brown method towards sales forecasting system with a linear and non-stationary data trend 3

Double exponential smoothing brown method towards sales fore...

引用

3rd International Conference on Vocational Education and Technology, IConVET 2020

作者： Dharmawan, P.A.S. Indradewi, I.G.A.A.D. Study Program of Informatics Engineering Stmik Stikom Indonesia Denpasar-Bali80225 Indonesia

UD Parama Store is a trading company engaged in selling retail goods that sell various types of daily necessities retail. The major problems occurred are the difficulty in predicting the sales due to the maturity level of experience, customer demand changes, and the owner's limited memory. Therefore, there should be an increase of merchandise stock to prevent any sudden decrease in sales and overcome the stock shortage when there is an increase in sales. In this current research, a sales forecasting web-based system was designed and built to assist the owner in predicting the number of sales in the next period. As a result, decisions can be made in determining the number of goods to be provided. The forecasting method used was double exponential smoothing brown by improving forecasting, averaging (smoothing) the past value of a time coherent data, and decreasing (exponential), which requires one parameter only. It was used to increase and decrease the linear and non-stationary data. The calculation of forecasting accuracy using the MAPE (Mean Absolute Percentage Error) method in forecasting sales of merchandise produces the smallest error rate ranging from 7.99% to 32.42% for 10 different items. © Published under licence by IOP Publishing Ltd.

关键词： Sales

来源：评论

学校读者我要写书评

暂无评论

Modified genetic algorithm for employee work shifts scheduling optimization 3

Modified genetic algorithm for employee work shifts scheduli...

引用

3rd International Conference on Vocational Education and Technology, IConVET 2020

作者： Saraswati, N.W.S. Artakusuma, I.D.M.D. Indradewi, I.G.A.A.D. Study Program of Informatics Engineering Stmik Stikom Indonesia Denpasar-Bali80225 Indonesia

Arranging an employee shift work's schedule requires high accuracy. It is because we have to pay attention to several constraints simultaneously. The genetic algorithm presents as a method which can automatize the process of arranging the schedule as well as optimizing the result of the schedule. The Shala Bali is a hospitality business whose scheduling was complicated because of the number of the employes. This research aimed at producing a shift work schedule of the employees in a week and to know the optimum genetic algorithm parameter in this case. The constraints that were taken into account in the arrangement of the schedule included the schedule conflict of the employees in one shift, schedule conflict of employees in 1 day, the same composition of employees per shift, employees should not get morning shifts after being in night shift the night before, each shift has at least 1 employee in the front office, and each employee is required to get 1 day off within 1 schedule period. This study was able to produce an optimal work schedule of employees with crossover probability (Pc) of 0.6, and mutation probability (Pm) of 0.3. The modification algorithm in chromosome generation and chromosome structure in this study results that changes in gene length (additional number of employees) do not have to be followed by an increase in the number of chromosome populations to get optimum results. © Published under licence by IOP Publishing Ltd.

关键词： Chromosomes

来源：评论

学校读者我要写书评

暂无评论

Forecasting system using single exponential smoothing with golden section optimization 2

Forecasting system using single exponential smoothing with g...

引用

2nd International Conference on Vocational Education and Technology, IConVET 2019

作者： Widitriani, N.P.S. Parwita, W.G.S. Meinarni, N.P.S. Study Program of Informatics Engineering STMIK STIKOM Indonesia Denpasar-Bali80225 Indonesia

Puri Wira Mahkota Ltd is a distributor company in the automotive parts industry. Puri Wira Mahkota has an obstacles in preparing weekly order on stock to the primary production unit. They predict total sales of product in the coming period just by using an expectation from stock administration staff then it will be compared with sales data in the previous period without any calculations using a definitive formula. Forecasting calculation could support Puri Wira Mahkota to predict the quantity of product. The purpose of this research is to apply the calculation process that can predict the quantity of product should be produced by the primary production unit every month and also to prevent over stock and out of stock. The method used in the forecasting calculation is Single Exponential Smoothing. To optimize Single Exponential Smoothing we used the Golden Section. The principle of the Golden Section is to reduce the alpha area boundary so it will produce an ideal forecast value with the minimum MAPE (Mean Absolute Percentage Error) level. This research shows the results that the Golden Section finds the optimal forecasting value with a level of MAPE (Mean Absolute Percentage Error) of 43.39%. © 2020 IOP Publishing Ltd. All rights reserved.

关键词： Sales

来源：评论

学校读者我要写书评

暂无评论

Design of IOT based facial recognition door access control locker services at tourist attractions in Bali 2

Design of IOT based facial recognition door access control l...

引用

2nd International Conference on Vocational Education and Technology, IConVET 2019

作者： Widiartha, K.K. Ekayana, A.A.G. Study Program of Informatics Engineering STMIK STIKOM Indonesia Denpasar-Bali80225 Indonesia

Safety and comfort are things that everyone wants to get on a trip. Every tourist who wants a vacation to one of the tourist destinations will prioritize these two things. Leisure in travel is influenced by the availability of supporting facilities. From accommodation, restaurants to facilities provided by tourism actors in each tourism object. One problem is the lack of facilities in attractions such as luggage storage (lockers), many tourists in enjoying tourism objects carrying luggage such as backpacks;groceries to backpacks for tourists are classified as backpackers. Of course this will disrupt the comfort of traveling. In the research conducted, researchers wanted to design a system of goods storage services as a form of supporting industry 4.0 in the tourism environment. By designing an image of the storage media of goods by implementing access control for doors in each locker based on image processing supported by a microcontroller. The method used that focuses on facial recognition as image input using the PCA (Principal Component Analysis) method, which can measure the Euclidean distance in an object's face image. The design as a blueprint makes it easier to develop the next item of storage media (lockers).The results of this study indicate that the facial recognition method used is able to extract facial features and be recalibrated to activate the locker that is owned by the face object. © 2020 IOP Publishing Ltd. All rights reserved.

关键词： Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

Fuzzy Time Series for Forecasting Railway Passengers in Indonesia 4

Fuzzy Time Series for Forecasting Railway Passengers in Indo...

引用

4th International Conference on Electrical, Electronics, informatics, and Vocational Education, ICE-ELINVO 2021

作者： Fatih Rizqon, Muhammad Jati, Handaru Department of Electronics and Informatics Engineering Education Postgraduate Program Yogyakarta State University Indonesia

Some fuzzy time series models have their own advantages and disadvantages. In addition, these models sometimes are complex and claimed to have better forecasting result than each other. The suitable model for forecasting depends on a wide variety of considerations. The models proposed by Chen (1996) applied simplified arithmetic operations and claimed more efficiency than before. The model proposed by Chen was introduced in 1996 and still exists in several previous studies. This research aims to forecast the number of railway passengers in Indonesia using the fuzzy time series. In addition, this research also evaluates the forecasting results based on mean absolute error (MAE) and mean absolute percentage error (MAPE). The results showed the forecasting results in this research has accuracy for 86.6%. © 2021 Institute of Physics Publishing. All rights reserved.

关键词： Forecasting

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：