Spiking neural networks (SNNs), inspired by the biological neural processing of the brain, are vastly growing due to their higher potential to handle spatiotemporal patterns with lower energy consumption, especially, ...
详细信息
ISBN:
(纸本)9798350350494;9798350350500
Spiking neural networks (SNNs), inspired by the biological neural processing of the brain, are vastly growing due to their higher potential to handle spatiotemporal patterns with lower energy consumption, especially, if implemented on neuromorphic devices. In this study, we propose self-supervised contrastive learning (SSL) for SNNs to learn informative latent representations from a large set of unlabeled data. The proposed SSL pre-trained SNN is then fine-tuned on a small set of labeled samples of a downstream supervised task. To evaluate the proposed method, we trained convolutional SNNs using SSL on MNIST and CIFAR10 datasets with 80% of images as unlabeled samples, then fine-tuned the networks on the remaining 20% images. The proposed SSL-based SNNs could reach 94.23% and 62.24% recognition accuracies on testing sets of MNIST and CIFAR10, respectively.
The proceedings contain 18 papers. The special focus in this conference is on Information Technologies and Intelligent Decision Making Systems. The topics include: Comparative Analysis of Traditional machinelearning ...
ISBN:
(纸本)9783031603174
The proceedings contain 18 papers. The special focus in this conference is on Information Technologies and Intelligent Decision Making Systems. The topics include: Comparative Analysis of Traditional machinelearning Approaches for Time Series Clustering Under Colored Noise;on the Open Transport data Analysis Platform;investigation of the Characteristics of a Frequency Diversity Array Antenna;comparative Analysis of Fuzzy Controllers in a Truck Cruise Control System;implementation of a Blockchain-Based Software Tool to Verify the Authenticity of Paper Documents;development of Methods and Algorithms for Dimension Reduction of Space Description for patternrecognition Problem;service for Checking Students’ Written Work Using a Neural Network;implementing a Jenkins Plugin to Visualize Continuous Integration Pipelines;elimination of Optical Distortions Arising from In Vivo Investigation of the Mouse Brain;quantum Fourier Transform in Image Processing;choosing an Information Protection Mechanism Based on the Discrete Programming Method;application of machinelearning Methods for Annotating Boundaries of Meshes of Perineuronal Nets;diagnostics of Animals Diseases Based on the Principles of Neutrosophic Sets and Sugeno Fuzzy Inference;the Technique of Processing Non-Gaussian data Based on Artificial Intelligence;development of Automation and Control System of Waste Gas Production Process Based on Information Technology;machinelearning and datamining.
Traditional economic growth forecasting methods, though somewhat successful, have limitations such as assuming a linear data generation process and ignoring the nonlinear relationships between economic variables. This...
详细信息
machinelearning (ML) is one of the effective security approaches to build cyber-attacks detection systems in Wireless Sensor Networks (WSNs). ML leverages the power of data analysis and patternrecognition to detect ...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
machinelearning (ML) is one of the effective security approaches to build cyber-attacks detection systems in Wireless Sensor Networks (WSNs). ML leverages the power of data analysis and patternrecognition to detect and classify various types of cyber-attacks to enhance the security of WSNs. A well-constructed dataset is one of the key factors that significantly impact the performance and generalization capabilities of any ML classifier trained on it. In this paper, we evaluate the effectiveness of two datasets: WSN-DS and WSN-BFSF which are specialized for Denial-of-Service (DoS) attacks targeting WSNs. We compare the two datasets in terms of their key characteristics, dataset quality, and ML classification performance. Mutual Information (MI) and Recursive Feature Elimination (RFE) are used for feature selection. The dataset quality is measured using statistical information calculation. The ML classification performance is investigated for three supervised ensemble techniques: LightGBM, bagging, and stacking using evaluation metrics including probability of detection, probability of false alarm, probability of misdetection, classification accuracy, model size, and processing time.
data analysis and mining play an important role in the research of intelligent information management system, but there is a problem of inaccurate information management. Traditional machinelearning cannot solve the ...
详细信息
The healthcare industry produces a huge amount of complex data about patient records. This data can be processed and mined to uncover hidden patterns, which will provide professionals in the healthcare field with addi...
详细信息
The proceedings contain 299 papers. The topics discussed include: time series modelling approach for predictive analytics;research on the recognition of accounting information distortion by random forest algorithm;rec...
ISBN:
(纸本)9798350318609
The proceedings contain 299 papers. The topics discussed include: time series modelling approach for predictive analytics;research on the recognition of accounting information distortion by random forest algorithm;reconstruction and update of 3D model of mechanical products based on 3D point cloud data;prediction of soil organic carbon content using machinelearning based fuzzy C-means clustering;fault detection and classification in power transmission lines using discrete wavelet transform-based swish recurrent neural network;immersive dramatic space 3D layout using panoramic image reconstruction algorithm;multi-objective optimization of noise in a high-speed railway by a hybrid algorithm;new energy vehicle customer mining model based on machinelearning algorithm;and robotization of agriculture using image processing techniques.
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
In the context of the rapid development of computer technology, people's lives have been inseparable from high-tech technology, which can bring convenience to people. In the field of computer vision, human activit...
详细信息
The proceedings contain 70 papers. The topics discussed include: research on the optimization of driving strategy of intelligent networked vehicles based on big data technology;a tennis momentum analysis method based ...
ISBN:
(纸本)9798400718120
The proceedings contain 70 papers. The topics discussed include: research on the optimization of driving strategy of intelligent networked vehicles based on big data technology;a tennis momentum analysis method based on Gaussian dynamics and machinelearning;sentiment and user churn analysis for online platform business using yelp studies;research on commercial vehicle replacement prediction model based on improved attention mechanism;cross-sectional industrialization factors’ contribution to heat wave risk classification;cross-domain person search data augmentation based on style transfer;recurrent convolutional fact graph neural networks for temporal knowledge graph reasoning;and research on network quality evaluation system based on the entropy weight method.
暂无评论