The proceedings contain 43 papers. The topics discussed include: application research of model-free reinforcement learning under the condition of conditional transfer function with coupling factors;expected regret min...
ISBN:
(纸本)9781450375511
The proceedings contain 43 papers. The topics discussed include: application research of model-free reinforcement learning under the condition of conditional transfer function with coupling factors;expected regret minimization for Bayesian optimization with student’s-t processes;a mining frequent itemsets algorithm in stream data based on sliding time decay window;experimental and theoretical scrutiny of the geometric derivation of the fundamental matrix;dual-precision deep neural network;annotating documents using active learning methods for a maintenance analysis application;offline handwritten Chinese character recognition based on improved Googlenet;a network combining local features and attention mechanisms for vehicle re-identification;and a spatial attention-enhanced multi-timescale graph convolutional network for skeleton-based action recognition.
The proceedings contain 39 papers. The topics discussed include: application research of model-free reinforcement learning under the condition of conditional transfer function with coupling factors;expected regret min...
ISBN:
(纸本)9781450375511
The proceedings contain 39 papers. The topics discussed include: application research of model-free reinforcement learning under the condition of conditional transfer function with coupling factors;expected regret minimization for Bayesian optimization with student’s-t processes;research on unbalanced data processing algorithm base Tomeklinks-Smote;a mining frequent itemsets algorithm in stream data based on sliding time decay window;experimental and theoretical scrutiny of the geometric derivation of the fundamental matrix;annotating documents using active learning methods for a maintenance analysis application;offline handwritten Chinese character recognition based on improved Googlenet;a network combining local features and attention mechanisms for vehicle re-identification;and a spatial attention-enhanced multi-timescale graph convolutional network for skeleton-based action recognition.
The proceedings contain 76 papers. The topics discussed include: ship detection with optical image based on attention and loss improved YOLO;ECG characteristic detection using DenseNet based on attention mechanism and...
ISBN:
(纸本)9781665499507
The proceedings contain 76 papers. The topics discussed include: ship detection with optical image based on attention and loss improved YOLO;ECG characteristic detection using DenseNet based on attention mechanism and feature pyramid;a method to detect the onsets and ends of paroxysmal atrial fibrillation episodes based on sliding window and coding;dynamic feature extraction using I-vector for video fire detection;bimodal information fusion network for salient object detection based on transformer;person re-identification method based on multi-view and attention mechanism;research on road unevenness recognition method based on off-road vehicle driving characteristic;and vehicle re-identification approach combining multiple attention mechanisms and style transfer.
This research paper presents a brief review of ten popular unsupervised algorithms widely utilized in patternrecognition publications. The algorithms are assessed based on their popularity, strengths, limitations, an...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
This research paper presents a brief review of ten popular unsupervised algorithms widely utilized in patternrecognition publications. The algorithms are assessed based on their popularity, strengths, limitations, and resource requirements. Considering these factors, we propose two most-preferred algorithms suitable for adoption in IDS (Intrusion Detection Systems) to address the problems associated with Zero Day exploits or attacks. Our review of the surveyed algorithms facilitated the recommendation of specific algorithms that can enhance IDS capabilities in detecting and mitigating Zero-Day attacks and anomalous intrusion attempts. These algorithms leverage unsupervised learning techniques to overcome the limitations of traditional signature-based approaches. By incorporating these algorithms, IDS can better handle sophisticated and evolving attacks that often evade detection. In conclusion, this research provides valuable insights into the strengths, limitations, and resource requirements of popular unsupervised algorithms used in patternrecognition. It highlights the potential of adopting these algorithms in IDS systems to bolster their ability to detect and respond to Zero-Day attacks. By recommending the integration of these algorithms, we contribute to the development of intelligent IDS solutions that can adapt to dynamic threat landscapes.
Implantable cardiac devices (ICDs) are often used as an effective treatment for arrhythmia. Although these devices have access to a live Electrocardiogram (ECG) stream, currently they do not offer on-device classifica...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
Implantable cardiac devices (ICDs) are often used as an effective treatment for arrhythmia. Although these devices have access to a live Electrocardiogram (ECG) stream, currently they do not offer on-device classification of arrhythmia due to the limited computing capability and severe power constraints. In this paper we propose a low-energy computing method for extracting shape-based features from ECG in combination with machine learning techniques for classifying nine different cases of arrhythmia. This is achieved by using a Localized Longest Common Subsequence (LLCS) algorithm which has low computational requirements that allows on-device execution. The proposed method strongly focuses on maintaining minimal energy and computational footprint, in line with the operating constraints of implantable devices. To demonstrate the energy efficiency and low computation load of the proposed method we implement the classification pipeline on a low-power RISC microcontroller and compare its performance with existing classification techniques. The classification accuracy and energy of the proposed method is compared with state-of-the art research in arrhythmia classification.
Explainability is needed to establish confidence in machine learning results. Some explainable methods take a post hoc approach to explain the weights of machine learning models, others highlight areas of the input co...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
Explainability is needed to establish confidence in machine learning results. Some explainable methods take a post hoc approach to explain the weights of machine learning models, others highlight areas of the input contributing to decisions. These methods do not adequately explain decisions, in plain terms. Explainable property-based systems have been shown to provide explanations in plain terms, however, they have not performed as well as leading unexplainable machine learning methods. This research focuses on the importance of metrics to explainability and contributes two methods yielding performance gains. The first method introduces a combination of explainable and unexplainable flows, proposing a metric to characterize explainability of a decision. The second method compares classic metrics for estimating the effectiveness of neural networks in the system, posing a new metric as the leading performer. Results from the new methods and examples from handwritten datasets are presented.
Food image recognition has drawn much attention recently because of its potential to transform the food business by automating food identification and streamlining the ordering and delivery process. One of the signifi...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
Food image recognition has drawn much attention recently because of its potential to transform the food business by automating food identification and streamlining the ordering and delivery process. One of the significant challenges in developing a food image recognition system is the high variability in the appearance of food items. The primary objective of this research is to build a food image recognition system based on deep learning. We propose two majority voting ensemble models, which outperform their state-of-the-art level-0 CNN models, EfficientNetV2S, InceptionV3, and Fusion - LSTM + InceptionV3. Furthermore, we propose seven meta-learner stacking models based on the three level-0 models. The best performing stacking model was the SVM linear stacking model and achieved 93.1% classification accuracy on the challenging Food-101 dataset. It improved upon the three level-0 models and the two voting ensemble models. MLP stacking model also improved upon the level-0 models and the voting ensemble models.
This document presents the 3nd internationalconference on Visual pattern Extraction and recognition for Cultural Heritage Understanding (VIPERC 2024), a premier forum for presenting academic and industry papers on bi...
详细信息
Electronic nose (e-nose) technology has become a powerful tool for identifying and evaluating complex scents in a variety of contexts, such as environmental monitoring, medical diagnostics, and food quality control. T...
详细信息
The proceedings contain 58 papers. The topics discussed include: development of a low cost IoT integrated system in precision smart garden;spectral and morphological classification of celestial objects using physics i...
ISBN:
(纸本)9798331534400
The proceedings contain 58 papers. The topics discussed include: development of a low cost IoT integrated system in precision smart garden;spectral and morphological classification of celestial objects using physics informed machine learning;development of an accelerometer-based data acquisition system for hand gesture recognition;lung cancer prediction and risk assessment: a machine learning approach integrating symptoms and etiological factors;a comparative analysis of short-term solar power forecasting using machine learning methods;intelligent fire surveillance: a deep learning framework for accurate indoor fire detection;robust phishing URL classification using FastText character embeddings and hybrid deep learning;and effective disease recognition in cucumbers: a web-based application using transfer learning models.
暂无评论