Due the unbalanced melanoma data and the complexity and resolution of the melanoma image backgrounds, classification of the melanoma regions is very challenging. In this paper, EffNet B5 models with different augmenta...
详细信息
Restricted Boltzmann machines (RBMs) is one of machinelearning's methods which within past decades, the development of RBMs has quite increase. Researches of RBMs focused on theories and applications of RBMs. The...
详细信息
ISBN:
(纸本)9781450366427
Restricted Boltzmann machines (RBMs) is one of machinelearning's methods which within past decades, the development of RBMs has quite increase. Researches of RBMs focused on theories and applications of RBMs. The application of RBMs has proofed that RBMs good at finishing many tasks, such as feature extraction method, document modeling, representation learning, classification and others. The RBMs' theories also have great movements, such as the development of the learning algorithm and inference techniques of RBMs. The key factors making the RBM success on finishing task are the learning algorithm and inference techniques. They motivated the development of inference techniques which successfully improved the deep neural network (DNN) performance. The aim of this research is reviewing the various types of RBMs as the application side, and the development of learning algorithm and inference techniques as theoretical side. Hopefully, it could motivate more development on the RBMs in order to contribute on overcoming implementation tasks especially on image processing tasks.
In modern sonar systems, automatic recognition of underwater targets has always been one of the key technologies in research. In recent years, classification and recognition methods based on machinelearning have been...
详细信息
ISBN:
(纸本)9781728151021
In modern sonar systems, automatic recognition of underwater targets has always been one of the key technologies in research. In recent years, classification and recognition methods based on machinelearning have been widely used in underwater acoustic field where good results have been achieved. Compared with monostatic active sonar, multi-static active sonar can simultaneously acquire the forward, lateral, and backscattering information of the target, and can obtain more accurate and stable target recognition result. Furthermore, the transmit waveform of active sonar effort the performance in complex ocean environment. As all known, the signals transmitted by cetaceans have the characteristics of strong anti-jamming ability, high positioning accuracy, et al. Accordingly, the performance of multi-static active sonar target recognition based on bionic signal is investigated in this paper. Besides, the machinelearning methods are applied to the recognition of echo signals, so that further good results and conclusions are obtained.
As infrasonic signals can through objects and propagate at a long distance, infrasound sensors are widely applied in wireless sensor networks to monitor environment events of a large area. The signal conditions are us...
详细信息
ISBN:
(纸本)9781450364027
As infrasonic signals can through objects and propagate at a long distance, infrasound sensors are widely applied in wireless sensor networks to monitor environment events of a large area. The signal conditions are usually complex and have various characteristics while monitoring the large area. Different features in both time and frequency domains should be extracted and considered. Big data increases the computation complexity, and the wrong selection of features may decreases the accuracy in event prediction. To overcome this problem, a query-based learning method is applied to select the proper features for smart edge computing in machinelearning. Experimental results show that the proposed method provides good performance when comparing with previous feature selection methods.
Surfing internet becomes common now-a-days that gave a chance for intruders to steal information. Therefore security is very important to detect any unwanted activities by using intrusion detection system. Intrusion d...
详细信息
Feature representation and feature fusion are important factors in image classification problem. In this paper, the local features, mid-level features and convolutional features are combined using the multiple kernel ...
详细信息
ISBN:
(纸本)9781728151021
Feature representation and feature fusion are important factors in image classification problem. In this paper, the local features, mid-level features and convolutional features are combined using the multiple kernel learning method. Experimental results show that the local features, mid-level features and convolutional features can be fused effectively to improve the classification performance about 4%-6% on several popular benchmarks.
in recent years, multi-label classification problem has become a controversial issue. In this kind of classification, each sample is associated with a set of class labels. Ensemble approaches are supervised learning a...
详细信息
ISBN:
(纸本)9781509058204
in recent years, multi-label classification problem has become a controversial issue. In this kind of classification, each sample is associated with a set of class labels. Ensemble approaches are supervised learning algorithms in which an operator takes a number of learning algorithms, namely base-level algorithms and combines their outcomes to make an estimation. The simplest form of ensemble learning is to train the base-level algorithms on random subsets of data and then let them vote for the most popular classifications or average the predictions of the base-level algorithms. In this study, an ensemble learning method is proposed for improving multi-label classification evaluation criteria. We have compared our method with well-known base-level algorithms on some data sets. Experiment results show the proposed approach outperforms the base well-known classifiers for the multi-label classification problem.
The spread of the Internet and mobile devices has made it easier, faster and more widely to disseminate information. But rumors also spread quickly through the Internet, which can have a big impact on people's liv...
详细信息
In this article, we designed an automatic Chinese text classification system aiming to implement a system for classifying news texts. We propose two improved classification algorithms as two different choices for user...
详细信息
ISBN:
(纸本)9781728133232
In this article, we designed an automatic Chinese text classification system aiming to implement a system for classifying news texts. We propose two improved classification algorithms as two different choices for users to choose and then our system uses the chosen method for the obtaining of the classified result of the input text. There are two improved algorithms, one is k-Bayes using hierarchy conception based on NB method in machinelearning field and another one adds attention layer to the convolutional neural network in deep learning field. Through experiments, our results showed that improved classification algorithms had better accuracy than based algorithms and our system is useful for making classifying news texts more reasonably and effectively.
The proceedings contain 15 papers. The topics discussed include: identifying optimal features for multi-channel acoustic scene classification;analysis of CNN architectures for pose estimation of noisy 3-D face images;...
ISBN:
(纸本)9781728138732
The proceedings contain 15 papers. The topics discussed include: identifying optimal features for multi-channel acoustic scene classification;analysis of CNN architectures for pose estimation of noisy 3-D face images;meteorite hunting using deep learning and UAVs;real-time dynamic security for ProSe in 5G;image steganography using YCbCr color space and matrix pattern;cost effective real time vision interface for off line simulation of Fanuc robots;smart healthcare systems on improving the efficiency of healthcare services;time-domain color mapping for color vision deficiency assistive technology;and spatio-temporal analysis andmachinelearning for traffic accidents prediction.
暂无评论