scholarly journals Completely Automated CNN Architecture Design Based on Blocks

2020 ◽  
Author(s):  
Y Sun ◽  
Bing Xue ◽  
Mengjie Zhang ◽  
GG Yen
Keyword(s):  
Genetic Algorithm ◽  
Classification Accuracy ◽  
Domain Knowledge ◽  
State Of The Art ◽  
Classification Performance ◽  
Data Sets ◽  
Computational Resource ◽  
Problem Domain ◽  
Benchmark Data ◽  
Interested User

© 2019 IEEE. The performance of convolutional neural networks (CNNs) highly relies on their architectures. In order to design a CNN with promising performance, extensive expertise in both CNNs and the investigated problem domain is required, which is not necessarily available to every interested user. To address this problem, we propose to automatically evolve CNN architectures by using a genetic algorithm (GA) based on ResNet and DenseNet blocks. The proposed algorithm is completely automatic in designing CNN architectures. In particular, neither preprocessing before it starts nor postprocessing in terms of CNNs is needed. Furthermore, the proposed algorithm does not require users with domain knowledge on CNNs, the investigated problem, or even GAs. The proposed algorithm is evaluated on the CIFAR10 and CIFAR100 benchmark data sets against 18 state-of-the-art peer competitors. Experimental results show that the proposed algorithm outperforms the state-of-the-art CNNs hand-crafted and the CNNs designed by automatic peer competitors in terms of the classification performance and achieves a competitive classification accuracy against semiautomatic peer competitors. In addition, the proposed algorithm consumes much less computational resource than most peer competitors in finding the best CNN architectures.

scholarly journals Completely Automated CNN Architecture Design Based on Blocks

2020 ◽  
Author(s):  
Y Sun ◽  
Bing Xue ◽  
Mengjie Zhang ◽  
GG Yen
Keyword(s):  
Genetic Algorithm ◽  
Classification Accuracy ◽  
Domain Knowledge ◽  
State Of The Art ◽  
Classification Performance ◽  
Data Sets ◽  
Computational Resource ◽  
Problem Domain ◽  
Benchmark Data ◽  
Interested User

© 2019 IEEE. The performance of convolutional neural networks (CNNs) highly relies on their architectures. In order to design a CNN with promising performance, extensive expertise in both CNNs and the investigated problem domain is required, which is not necessarily available to every interested user. To address this problem, we propose to automatically evolve CNN architectures by using a genetic algorithm (GA) based on ResNet and DenseNet blocks. The proposed algorithm is completely automatic in designing CNN architectures. In particular, neither preprocessing before it starts nor postprocessing in terms of CNNs is needed. Furthermore, the proposed algorithm does not require users with domain knowledge on CNNs, the investigated problem, or even GAs. The proposed algorithm is evaluated on the CIFAR10 and CIFAR100 benchmark data sets against 18 state-of-the-art peer competitors. Experimental results show that the proposed algorithm outperforms the state-of-the-art CNNs hand-crafted and the CNNs designed by automatic peer competitors in terms of the classification performance and achieves a competitive classification accuracy against semiautomatic peer competitors. In addition, the proposed algorithm consumes much less computational resource than most peer competitors in finding the best CNN architectures.

scholarly journals A Multi-Branch Feature Fusion Strategy Based on an Attention Mechanism for Remote Sensing Image Scene Classification

2021 ◽  
Vol 13 (10) ◽  
pp. 1950
Author(s):  
Cuiping Shi ◽  
Xin Zhao ◽  
Liguo Wang
Keyword(s):  
Remote Sensing ◽  
Feature Extraction ◽  
Classification Accuracy ◽  
Feature Fusion ◽  
State Of The Art ◽  
Rapid Development ◽  
Remote Sensing Image ◽  
Classification Performance ◽  
Attention Mechanism ◽  
Scene Classification

In recent years, with the rapid development of computer vision, increasing attention has been paid to remote sensing image scene classification. To improve the classification performance, many studies have increased the depth of convolutional neural networks (CNNs) and expanded the width of the network to extract more deep features, thereby increasing the complexity of the model. To solve this problem, in this paper, we propose a lightweight convolutional neural network based on attention-oriented multi-branch feature fusion (AMB-CNN) for remote sensing image scene classification. Firstly, we propose two convolution combination modules for feature extraction, through which the deep features of images can be fully extracted with multi convolution cooperation. Then, the weights of the feature are calculated, and the extracted deep features are sent to the attention mechanism for further feature extraction. Next, all of the extracted features are fused by multiple branches. Finally, depth separable convolution and asymmetric convolution are implemented to greatly reduce the number of parameters. The experimental results show that, compared with some state-of-the-art methods, the proposed method still has a great advantage in classification accuracy with very few parameters.

scholarly journals Emotion Assessment Using Feature Fusion and Decision Fusion Classification Based on Physiological Data: Are We There Yet?

Sensors ◽  
2020 ◽  
Vol 20 (17) ◽  
pp. 4723
Author(s):  
Patrícia Bota ◽  
Chen Wang ◽  
Ana Fred ◽  
Hugo Silva
Keyword(s):  
Emotion Recognition ◽  
Classification Accuracy ◽  
Feature Fusion ◽  
State Of The Art ◽  
Electrodermal Activity ◽  
Decision Fusion ◽  
Classification Performance ◽  
Physiological Data ◽  
Systematic Analysis ◽  
Public Datasets

Emotion recognition based on physiological data classification has been a topic of increasingly growing interest for more than a decade. However, there is a lack of systematic analysis in literature regarding the selection of classifiers to use, sensor modalities, features and range of expected accuracy, just to name a few limitations. In this work, we evaluate emotion in terms of low/high arousal and valence classification through Supervised Learning (SL), Decision Fusion (DF) and Feature Fusion (FF) techniques using multimodal physiological data, namely, Electrocardiography (ECG), Electrodermal Activity (EDA), Respiration (RESP), or Blood Volume Pulse (BVP). The main contribution of our work is a systematic study across five public datasets commonly used in the Emotion Recognition (ER) state-of-the-art, namely: (1) Classification performance analysis of ER benchmarking datasets in the arousal/valence space; (2) Summarising the ranges of the classification accuracy reported across the existing literature; (3) Characterising the results for diverse classifiers, sensor modalities and feature set combinations for ER using accuracy and F1-score; (4) Exploration of an extended feature set for each modality; (5) Systematic analysis of multimodal classification in DF and FF approaches. The experimental results showed that FF is the most competitive technique in terms of classification accuracy and computational complexity. We obtain superior or comparable results to those reported in the state-of-the-art for the selected datasets.

Multi-Frequency Matched-Field Inversion of Benchmark Data Using a Genetic Algorithm

1998 ◽  
Vol 06 (01n02) ◽  
pp. 135-150 ◽  
Author(s):  
D. G. Simons ◽  
M. Snellen
Keyword(s):  
Genetic Algorithm ◽  
Test Cases ◽  
Data Sets ◽  
True Parameter ◽  
Benchmark Data ◽  
Line Array ◽  
Standard Normal ◽  
Water Test ◽  
Parameter Values ◽  
Field Inversion

For a selected number of shallow water test cases of the 1997 Geoacoustic Inversion Workshop we have applied Matched-Field Inversion to determine the geoacoustic and geometric (source location, water depth) parameters. A genetic algorithm has been applied for performing the optimization, whereas the replica fields have been calculated using a standard normal-mode model. The energy function to be optimized is based on the incoherent multi-frequency Bartlett processor. We have used the data sets provided at a few frequencies in the band 25–500 Hz for a vertical line array positioned at 5 km from the source. A comparison between the inverted and true parameter values is made.

scholarly journals Pattern and Feature Selection by Genetic Algorithms in Nearest Neighbor Classification

2000 ◽  
Vol 4 (2) ◽  
pp. 138-145 ◽  
Author(s):  
Hisao Ishibuchi ◽  
◽  
Tomoharu Nakashima
Keyword(s):  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Feature Selection ◽  
Computer Simulations ◽  
Nearest Neighbor ◽  
Classification Performance ◽  
Data Sets ◽  
Nearest Neighbor Classification ◽  
Reference Set ◽  
Neighbor Classification

This paper proposes a genetic-algorithm-based approach for finding a compact reference set in nearest neighbor classification. The reference set is designed by selecting a small number of reference patterns from a large number of training patterns using a genetic algorithm. The genetic algorithm also removes unnecessary features. The reference set in our nearest neighbor classification consists of selected patterns with selected features. A binary string is used for representing the inclusion (or exclusion) of each pattern and feature in the reference set. Our goal is to minimize the number of selected patterns, to minimize the number of selected features, and to maximize the classification performance of the reference set. Computer simulations on commonly used data sets examine the effectiveness of our approach.

Joint Concept Correlation and Feature-Concept Relevance Learning for Multilabel Classification

2018 ◽  
Vol 30 (2) ◽  
pp. 526-545
Author(s):  
Xiaowei Zhao ◽  
Zhigang Ma ◽  
Zhi Li ◽  
Zhihui Li
Keyword(s):  
State Of The Art ◽  
Relational Learning ◽  
Classification Performance ◽  
Data Sets ◽  
Classification Methods ◽  
Average Precision ◽  
Multilabel Classification ◽  
Multimedia Annotation ◽  
Multilabel Learning ◽  
Significant Attention

In recent years, multilabel classification has attracted significant attention in multimedia annotation. However, most of the multilabel classification methods focus only on the inherent correlations existing among multiple labels and concepts and ignore the relevance between features and the target concepts. To obtain more robust multilabel classification results, we propose a new multilabel classification method aiming to capture the correlations among multiple concepts by leveraging hypergraph that is proved to be beneficial for relational learning. Moreover, we consider mining feature-concept relevance, which is often overlooked by many multilabel learning algorithms. To better show the feature-concept relevance, we impose a sparsity constraint on the proposed method. We compare the proposed method with several other multilabel classification methods and evaluate the classification performance by mean average precision on several data sets. The experimental results show that the proposed method outperforms the state-of-the-art methods.

An Integrated Framework for Generalized Nearest Prototype Classifier Design

1998 ◽  
Vol 06 (05) ◽  
pp. 437-457 ◽  
Author(s):  
Ludmila I. Kuncheva ◽  
James C. Bezdek
Keyword(s):  
Nearest Neighbor ◽  
Classification Performance ◽  
Data Sets ◽  
Viable Option ◽  
Classifier Design ◽  
Benchmark Data ◽  
Label Information ◽  
Nearest Neighbor Rule ◽  
Common Framework ◽  
Class Labels

We propose a Generalized Nearest Prototype Classifier (GNPC) as a common framework for a number of classification techniques. Specifically we consider clustering-and-relabeling; Parzen's classifier; radial basis functions (RBF) networks; learning vector quantization (LVQ) type classifiers; and nearest neighbor rules. To classify an unlabeled point x the GNPC combines the degrees of similarity of x to a set of prototypes. Five questions are addressed for these GNPC families: (1) How many prototypes do we need? (2) How are the prototypes found? (3) How are their class labels obtained? (4) How are the similarities defined? (5) How are the similarities and label information combined? The classification performance of a set of GNPCs is illustrated on two benchmark data sets: IRIS and the 2-spirals data. We study the resubstitution error of the GNPC as a function of the number of prototypes. Our conclusions are that: (a) unsupervised selection (or extraction) of prototypes followed by relabeling is inferior to the techniques that use labels to guide them towards prototypes; (b) the edited nearest neighbor rule is a viable option for GNPC design which has not received the attention it deserves.

Multi-Attribution Classification Based on Lower Integral with Genetic Algorithm

2014 ◽  
Vol 668-669 ◽  
pp. 1090-1093
Author(s):  
Ai Xia Chen ◽  
Jun Hua Li
Keyword(s):  
Genetic Algorithm ◽  
Linear Programming ◽  
Linear Programming Problem ◽  
Fuzzy Integral ◽  
Data Sets ◽  
Fuzzy Measures ◽  
Benchmark Data ◽  
Minimum Potential ◽  
And Performance ◽  
Set Function

Fuzzy integral has been widely used in multi-attribution classification when the interactions exist between the attributions. Because the fuzzy measure defined on the attributions represents the weights of all the attributions and the interactions between them. The lower integral is a type of fuzzy integral with respect to fuzzy measures, which represents the minimum potential of efficiency for a group of attributions with interaction. The value of lower integrals can be evaluated through solving a linear programming problem. Considering the lower integral as a classifier, this paper investigates its implementation and performance. The difficult step in the implementation is how to learn the non-additive set function used in lower integrals. And Genetic algorithm is used to solve the problem. Finally, numerical simulations on some benchmark data sets are given.

MULTISTAGE MUTUAL INFORMATION FOR INFORMATIVE GENE SELECTION

2011 ◽  
Vol 19 (04) ◽  
pp. 725-746 ◽  
Author(s):  
P. GANESH KUMAR ◽  
T. ARULDOSS ALBERT VICTOIRE
Keyword(s):  
Mutual Information ◽  
Classification Accuracy ◽  
Gene Selection ◽  
Classification Performance ◽  
Microarray Data Analysis ◽  
Data Sets ◽  
Criterion Function ◽  
Gene Expressions ◽  
Target Disease ◽  
Artificial Neural Network Ann

An important issue in the design of gene selection algorithm for microarray data analysis is the formation of suitable criterion function for measuring the relevance between different gene expressions. Mutual information (MI) is a widely used criterion function but it calculates the relevance on the entire samples only once which cannot exactly identify the informative genes. This paper proposes a novel idea of computing MI in stages. The proposed multistage mutual information (MSMI) computes MI, initially using all the samples and based on the classification performance produced by artificial neural network (ANN), MI is repeatedly calculated using only the unclassified samples until there is no improvement in the classification accuracy. The performance of the proposed approach is evaluated using ten gene expression data sets. Simulation result shows that the proposed approach helps to improve the discriminate power of the genes with regard to the target disease of a microarray sample. Statistical analysis of the test result shows that the proposed method selects highly informative genes and produces comparable classification accuracy than the other approaches reported in the literature.

scholarly journals MolRep: A Deep Representation Learning Library for Molecular Property Prediction

2021 ◽  
Author(s):  
Jiahua Rao ◽  
Shuangjia Zheng ◽  
Ying Song ◽  
Jianwen Chen ◽  
Chengtao Li ◽  
...  
Keyword(s):  
State Of The Art ◽  
Source Code ◽  
Representation Learning ◽  
Supplementary Information ◽  
Data Sets ◽  
Supplementary Data ◽  
Property Prediction ◽  
Average Rank ◽  
Benchmark Data ◽  
Classification Tasks

AbstractSummaryRecently, novel representation learning algorithms have shown potential for predicting molecular properties. However, unified frameworks have not yet emerged for fairly measuring algorithmic progress, and experimental procedures of different representation models often lack rigorousness and are hardly reproducible. Herein, we have developed MolRep by unifying 16 state-of-the-art models across 4 popular molecular representations for application and comparison. Furthermore, we ran more than 12.5 million experiments to optimize hyperparameters for each method on 12 common benchmark data sets. As a result, CMPNN achieves the best results ranked the 1st in 5 out of 12 tasks with an average rank of 1.75. Relatively, ECC has good performance in classification tasks and MAT good for regression (both ranked 1st for 3 tasks) with an average rank of 2.71 and 2.6, respectively.AvailabilityThe source code is available at: https://github.com/biomed-AI/MolRepSupplementary informationSupplementary data are available online.

Sign in / Sign up

Export Citation Format

Share Document