Efficient Continuous kNN Join over Dynamic High-dimensional Data DOI Creative Commons
Nimish Ukey, Guangjian Zhang, Zhengyi Yang

et al.

Research Square (Research Square), Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 18, 2023

Abstract Given a user dataset U and an object I, kNN join query in high-dimensional space returns the k nearest neighbors of each from I. The is basic necessary operation many applications, such as databases, data mining, computer vision, multi-media, machine learning, recommenda-tion systems, more. In real world, datasets frequently update dynamically objects are added or removed. this paper, we propose novel methods continuous over dynamic data. We firstly HDR+ Tree, which supports more efficient insertion, deletion, batch update. Further observed that existing rely on globally correlated for effec-tive dimensionality reduction, then HDR Forest. It clusters constructs multiple Trees to capture local correlations among As result, our Forest able process non-globally efficiently. Two optimisations applied proposed Forest, including precomputation PCA states items pruning-based recomputation during item deletion. For completeness work, also present proof computing distances reduced dimensions Tree. Extensive experiments real-world show outperform baseline algorithms naive RkNN

Language: Английский

A Comprehensive Review on Plant-Based Medications and Chemical Approaches for Autism Spectrum Disorders (ASDs) Psychopharmacotherapy DOI
Vrish Dhwaj Ashwlayan, Ratneshwar Kumar Ratnesh, Divya Sharma

et al.

Indian Journal of Microbiology, Journal Year: 2024, Volume and Issue: unknown

Published: April 10, 2024

Language: Английский

Citations

5
Attention-Focused Eye Gaze Analysis to Predict Autistic Traits Using Transfer Learning DOI Creative Commons
Ranjeet Vasant Bidwe, Sashikala Mishra,

Simi Bajaj

et al.

International Journal of Computational Intelligence Systems, Journal Year: 2024, Volume and Issue: 17(1)

Published: May 16, 2024

Abstract Autism spectrum disorder (ASD) is a complex developmental issue that affects the behavior and communication abilities of children. It extremely needed to perceive it at an early age. The research article focuses on attentiveness by considering eye positioning as key feature its implementation completed in two phases. In first phase, various transfer learning algorithms are implemented evaluated predict ASD traits available open-source image datasets Kaggle Zenodo. To reinforce result, fivefold cross-validation used dataset. Progressive pre-trained named VGG 16, 19, InceptionV3, ResNet152V2, DenseNet201, ConNextBase, EfficientNetB1, NasNetMobile, InceptionResNEtV2 establish correctness result. result being compiled analyzed ConvNextBase model has best diagnosing ability both datasets. This achieved prediction accuracy 80.4% with batch size rate 0.00002, 10 epochs 6 units, 80.71% Zenodo dataset 4, 4 units. found challenging nature compared existing model. Attentiveness parameter will accurately diagnose visual participant which helps automatic autistic traits. second phase proposed model, engrossed identifying uses dlib library HOG Linear SVM-based face detectors identify particular facial called EAR measure participants' based gaze analysis. If value less than 0.20 for more 100 consecutive frames, concludes un-attentive. generated special graph time period continuously plotting attention level. average depict participant.

Language: Английский

Citations

5
Design of the AI Application for the Detection of Autism Spectrum Disorder Using Deep Learning DOI

S. Seema,

Monica R. Mundada,

Meeradevi Meeradevi

et al.

Lecture notes in networks and systems, Journal Year: 2025, Volume and Issue: unknown, P. 577 - 586

Published: Jan. 1, 2025

Language: Английский

Citations

0
Bio-inspired swarm intelligence-based feature selection and classification for autism spectrum disorder detection DOI
N. M. Saravana Kumar,

Kannapiran Selvakumar,

V. Senthil Murugan

et al.

AIP conference proceedings, Journal Year: 2025, Volume and Issue: 3308, P. 030001 - 030001

Published: Jan. 1, 2025

Language: Английский

Citations

0
Evaluating the efficacy and site-specific performance of machine learning approaches: A comprehensive review of autism detection models DOI Creative Commons

Deblina Mazumder Setu,

Tania Islam,

Md Maklachur Rahman

et al.

Franklin Open, Journal Year: 2025, Volume and Issue: unknown, P. 100275 - 100275

Published: May 1, 2025

Language: Английский

Citations

0
Integrating routine blood biomarkers and artificial intelligence for supporting diagnosis of silicosis in engineered stone workers DOI Creative Commons
Daniel Morillo, Antonio León‐Jiménez, María Guerrero‐Chanivet

et al.

Bioengineering & Translational Medicine, Journal Year: 2024, Volume and Issue: 9(6)

Published: June 28, 2024

Abstract Engineered stone silicosis (ESS), primarily caused by inhaling respirable crystalline silica, poses a significant occupational health risk globally. ESS has no effective treatment and presents rapid progression from simple (SS) to progressive massive fibrosis (PMF), with respiratory failure death. Despite the use of diagnostic methods like chest x‐rays high‐resolution computed tomography, early detection remains challenging. Since routine blood tests have shown promise in detecting inflammatory markers associated disease, this study aims assess whether biomarkers, coupled machine learning techniques, can effectively differentiate between healthy individuals, subjects SS, PMF. To end, 107 men diagnosed silicosis, ex‐workers engineered (ES) sector, 22 male volunteers as controls not exposed ES dust were recruited. Twenty‐one primary biochemical derived peripheral extraction obtained retrospectively clinical hospital records. Relief‐ F features selection technique was applied, resulting subset 11 biomarkers used build five models, demonstrating high performance sensitivities specificities best case greater than 82% 89%, respectively. The percentage lymphocytes, angiotensin‐converting enzyme, lactate dehydrogenase indexes revealed, among others, cumulative importance for models. Our reveals that these could detect chronic status potentially serve supportive tool diagnosis, monitoring, silicosis.

Language: Английский

Citations

2
Linguistic summarization of visual attention and developmental functioning of young children with autism spectrum disorder DOI Creative Commons
Demet Öztürk, Sena Aydoğan, İbrahim Kök

et al.

Health Information Science and Systems, Journal Year: 2024, Volume and Issue: 12(1)

Published: July 16, 2024

Abstract Diagnosing autism spectrum disorder (ASD) in children poses significant challenges due to its complex nature and impact on social communication development. While numerous data analytics techniques have been proposed for ASD evaluation, the process remains time-consuming lacks clarity. Eye tracking (ET) has emerged as a valuable resource risk assessment, yet existing literature predominantly focuses predictive methods rather than descriptive that offer human-friendly insights. Interpretation of ET Bayley scales, widely used assessment tool, is challenging children. It should be understood clearly perform better analytic tasks screening. Therefore, this study addresses gap by employing linguistic summarization generate easily understandable summaries from raw scales. By integrating scores, aims improve identification with typically developing (TD). Notably, research represents one pioneering efforts linguistically summarize alongside presenting comparative results between TD. Through summarization, facilitates creation simple, natural language statements, offering first unique approach enhance screening contribute our understanding neurodevelopmental disorders.

Language: Английский

Citations

2
Efficient continuous kNN join over dynamic high-dimensional data DOI Creative Commons
Nimish Ukey, Guangjian Zhang, Zhengyi Yang

et al.

World Wide Web, Journal Year: 2023, Volume and Issue: 26(6), P. 3759 - 3794

Published: Sept. 11, 2023

Abstract Given a user dataset $$\varvec{U}$$ U and an object $$\varvec{I}$$ xmlns:mml="http://www.w3.org/1998/Math/MathML">I , kNN join query in high-dimensional space returns the $$\varvec{k}$$ xmlns:mml="http://www.w3.org/1998/Math/MathML">k nearest neighbors of each from . The is basic necessary operation many applications, such as databases, data mining, computer vision, multi-media, machine learning, recommendation systems, more. In real world, datasets frequently update dynamically objects are added or removed. this paper, we propose novel methods continuous over dynamic data. We firstly HDR $$^+$$ xmlns:mml="http://www.w3.org/1998/Math/MathML">+ Tree, which supports more efficient insertion, deletion, batch update. Further observed that existing rely on globally correlated for effective dimensionality reduction, then Forest. It clusters constructs multiple Trees to capture local correlations among As result, our Forest able process non-globally efficiently. Two optimisations applied proposed Forest, including precomputation PCA states items pruning-based recomputation during item deletion. For completeness work, also present proof computing distances reduced dimensions Tree. Extensive experiments real-world show outperform baseline algorithms naive RkNN

Language: Английский

Citations

2
Large data density peak clustering based on sparse auto-encoder and data space meshing via evidence probability distribution DOI Creative Commons
Lu Fang

ICST Transactions on Scalable Information Systems, Journal Year: 2024, Volume and Issue: 11

Published: Nov. 20, 2024

The development of big data analysis technology has brought new opportunities to the production and management various industries. Through mining in operation process enterprises by technology, internal associated even entire industry can be obtained. As a common method for large-scale statistical analysis, clustering effectively mine relationship within massive heterogeneous multidimensional data, complete unlabeled classification, provide support model data. Common density methods are time-consuming easy cause errors allocation, which affects accuracy clustering. Therefore we propose novel large peak based on sparse auto-encoder space meshing via evidence probability distribution. Firstly, deep learning is used achieve feature extraction dimensionality reduction input high-dimensional matrix through training. Secondly, meshed reduce calculation distance between sample points. When calculating local density, not only value grid itself, but also nearest neighbors considered, reduces influence subjective selection truncation results improves accuracy. threshold set ensure stability results. Using K-nearest neighbor information points, transfer distribution strategy proposed optimize remaining so as avoid joint error experimental show that algorithm higher better performance than other advanced algorithms artificial real sets.

Language: Английский

Citations

0
Efficient Continuous kNN Join over Dynamic High-dimensional Data DOI Creative Commons
Nimish Ukey, Guangjian Zhang, Zhengyi Yang

et al.

Research Square (Research Square), Journal Year: 2023, Volume and Issue: unknown

Published: Feb. 16, 2023

Abstract Given a user dataset U and an object I, kNN join query in high-dimensional space returns the k nearest neighbors of each U from I. The is basic and necessary operation many applications, such as databases, data mining, computer vision, multi-media, machine learning, recommendation systems, more. In real world, datasets frequently update dynamically objects are added or removed. this paper, we propose novel methods continuous over dynamic data. We firstly HDR+ Tree which supports more efficient insertion, deletion, batch update. Further observed that existing rely on globally correlated for effective dimen-sionality reduction, then HDR Forest. It clusters constructs multiple Trees to capture local correlations among As result, our Forest able process non-globally efficiently. Two optimisations applied proposed Forest, including precomputation PCA states items pruning-based recomputation during item deletion. For completeness work, also present proof computing distances reduced dimensions Tree. Extensive experiments real-world show outperform baseline algorithms naive RkNN

Language: Английский

Citations

0