scholarly journals Assessment of Deep Convolutional Neural Network Models for Species Identification of Forensically-Important Fly Maggots Based on Images of Posterior Spiracles

2022 ◽  
Author(s):  
Darlin Apasrawirote ◽  
Pharinya Boonchai ◽  
Paisarn Muneesawang ◽  
Wannacha Nakhonkam ◽  
Nophawan Bunchu
Keyword(s):  
Web Application ◽  
Confusion Matrix ◽  
Digital Camera ◽  
Network Models ◽  
Classification Model ◽  
Neural Network Models ◽  
Species Classification ◽  
Difficult Time ◽  
Novel Method ◽  
Level Identification

Abstract Forensic entomology is the branch of forensic science that is related to using arthropod specimens found in legal issues. Fly maggots are one of crucial pieces of evidence that can be used for estimating post-mortem intervals worldwide. However, the species-level identification of fly maggots is difficult, time consuming, and requires specialized taxonomic training. In this work, a novel method for the identification of different forensically-important fly species is proposed using convolutional neural networks (CNNs). The data used for the experiment were obtained from a digital camera connected to a compound microscope. We compared the performance of four widely used models that vary in complexity of architecture to evaluate tradeoffs in accuracy and speed for species classification including ResNet-101, Densenet161, Vgg19_bn, and AlexNet. In the validation step, all of the studied models provided 100% accuracy for identifying maggots of 4 species including Chrysomya megacephala (Diptera: Calliphoridae), Chrysomya (Achoetandrus) rufifacies (Diptera: Calliphoridae), Lucilia cuprina (Diptera: Calliphoridae), and Musca domestica (Diptera: Muscidae) based on images of posterior spiracles. However, AlexNet showed the fastest speed to process the identification model and presented a good balance between performance and speed. Therefore, the AlexNet model was selected for the testing step. The results of the confusion matrix of AlexNet showed that misclassification was found between C. megacephala and C. (Achoetandrus) rufifacies as well as between C. megacephala and L. cuprina. No misclassification was found for M. domestica. In addition, we created a web-application platform called thefly.ai to help users identify species of fly maggots in their own images using our classification model. The results from this study can be applied to identify further species by using other types of images. This model can also be used in the development of identification features in mobile applications. This study is a crucial step for integrating information from biology and AI-technology to develop a novel platform for use in forensic investigation.

scholarly journals Constructive Learning of Deep Neural Networks for Bigdata Analysis

2020 ◽  
Vol 9 (12) ◽  
pp. 311-322
Author(s):  
Soha Abd Mohamed El-Moamen ◽  
Marghany Hassan Mohamed ◽  
Mohammed F. Farghally
Keyword(s):  
Neural Network ◽  
Lung Cancer ◽  
Binary Classification ◽  
Network Models ◽  
Classification Model ◽  
Neural Network Models ◽  
Constructive Learning ◽  
The Neural Network ◽  
Rapid Pace ◽  
Better Than

The need for tracking and evaluation of patients in real-time has contributed to an increase in knowing people’s actions to enhance care facilities. Deep learning is good at both a rapid pace in collecting frameworks of big data healthcare and good predictions for detection the lung cancer early. In this paper, we proposed a constructive deep neural network with Apache Spark to classify images and levels of lung cancer. We developed a binary classification model using threshold technique classifying nodules to benign or malignant. At the proposed framework, the neural network models training, defined using the Keras API, is performed using BigDL in a distributed Spark clusters. The proposed algorithm has metrics AUC-0.9810, a misclassifying rate from which it has been shown that our suggested classifiers perform better than other classifiers.

scholarly journals Aspect-Level Sentiment Analysis Based on Position Features Using Multilevel Interactive Bidirectional GRU and Attention Mechanism

2020 ◽  
Vol 2020 ◽  
pp. 1-13
Author(s):  
Xiaodi Wang ◽  
Xiaoliang Chen ◽  
Mingwei Tang ◽  
Tian Yang ◽  
Zhen Wang
Keyword(s):  
Neural Network ◽  
Sentiment Analysis ◽  
Network Models ◽  
Attention Mechanism ◽  
Sentiment Classification ◽  
Classification Model ◽  
Neural Network Models ◽  
Position Information ◽  
Gated Recurrent Unit ◽  
Relative Position Information

The aim of aspect-level sentiment analysis is to identify the sentiment polarity of a given target term in sentences. Existing neural network models provide a useful account of how to judge the polarity. However, context relative position information for the target terms is adversely ignored under the limitation of training datasets. Considering position features between words into the models can improve the accuracy of sentiment classification. Hence, this study proposes an improved classification model by combining multilevel interactive bidirectional Gated Recurrent Unit (GRU), attention mechanisms, and position features (MI-biGRU). Firstly, the position features of words in a sentence are initialized to enrich word embedding. Secondly, the approach extracts the features of target terms and context by using a well-constructed multilevel interactive bidirectional neural network. Thirdly, an attention mechanism is introduced so that the model can pay greater attention to those words that are important for sentiment analysis. Finally, four classic sentiment classification datasets are used to deal with aspect-level tasks. Experimental results indicate that there is a correlation between the multilevel interactive attention network and the position features. MI-biGRU can obviously improve the performance of classification.

Double Multi-Head Attention-Based Capsule Network for Relation Classification

2021 ◽  
Author(s):  
Hongjun Heng ◽  
Renjie Li
Keyword(s):  
Neural Network ◽  
Feature Extraction ◽  
Language Processing ◽  
Layer Structure ◽  
Single Layer ◽  
Network Models ◽  
Classification Model ◽  
Neural Network Models ◽  
Comparable Performance ◽  
Relation Classification

Semantic relation classification is an important task in the field of nature language processing. The existing neural network relation classification models introduce attention mechanism to increase the importance of significant features, but part of these attention models only have one head which is not enough to capture more distinctive fine-grained features. Models based on RNN (Recurrent Neural Network) usually use single-layer structure and have limited feature extraction capability. Current RNN-based capsule networks have problem of improper handling of noise which increase complexity of network. Therefore, we propose a capsule network relation classification model based on double multi-head attention. In this model, we introduce an auxiliary BiGRU (Bidirectional Gated Recurrent Unit) to make up for the lack of feature extraction performance of single BiGRU, improve the bilinear attention through double multihead mechanism to enable the model to obtain more information of sentence from different representation subspace and instantiate capsules with sentence-level features to alleviate noise impact. Experiments on the SemEval-2010 Task 8 benchmark dataset show that our model outperforms most of previous state-of-the-art neural network models and achieves the comparable performance with F1 score of 85.3% in capsule network.

Prediction of Characteristics of Wave Breaking in Shallow Water Using Neural Network Techniques

2017 ◽  
Author(s):  
Nicholas Kouvaras ◽  
Manhar R. Dhanak
Keyword(s):  
Neural Network ◽  
Wave Height ◽  
Predictive Models ◽  
Wave Breaking ◽  
Laboratory Experiments ◽  
Network Models ◽  
Classification Model ◽  
Neural Network Models ◽  
Wave Setup ◽  
Fringing Reef

The characteristics of wave breaking over a fringing reef are considered using a set of laboratory experiments and the results are used to develop associated predictive models. Various methods are typically used to estimate the characteristics of nearshore wave breaking, mostly based on empirical, analytical and numerical techniques. Deo et al. (2003) used an artificial neural network approach to predict the breaking wave height and breaking depth for waves transforming over a range of simply sloped bottoms. The approach is based on using available representative data to train appropriate neural network models. The Deo et al. (2003) approach is extended here to predict other characteristics of wave breaking, including the type of wave breaking, and the position of breaking over a fringing reef, as well as the associated wave setup, and the rate of dissipation of wave energy, based on observations from a series of laboratory experiments involving monochromatic waves impacting on an idealized reef. Yao et al. (2013) showed that for such geometry, the critical parameter is the ratio of deep-water wave height to the depth of the shallow reef flat downstream of the position of wave breaking, H1/hs, rather than the slope of the reef. H1/hs, and the wave frequency parameter, fH1/g, are provided as inputs to the neural network models of the feed-forward type that are developed to predict the above characteristics of wave breaking. The models are trained using the experimental data. The breaker type classification model has a success rate of over 95%, implying that the neural networks method outperforms previously used criteria for classifying breaker types. The numeric prediction model for the dimensionless position of wave breaking also performs well, with a high degree of correlation between the predicted and actual positions of wave breaking. The performance is higher when only the plunging breaker instances are considered, but lower when only the spilling breaker instances are considered. The corresponding neural network models for wave setup within the surf zone and the difference in energy flux between the incident and broken wave have success rates of approximately 89% and 94% respectively. The method may be extended to provide predictive models for consideration of a range of natural coastal conditions, random waves, and various bottom profiles and complex geometry, based on training and testing of the models using representative field and laboratory observational data, in support of accurate prediction of near-shore wave phenomena.

scholarly journals A protection method of trained CNN model with a secret key from unauthorized access

2021 ◽  
Vol 10 ◽  
Author(s):  
AprilPyone Maungmaung ◽  
Hitoshi Kiya
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
State Of The Art ◽  
Network Models ◽  
The State ◽  
Secret Key ◽  
Neural Network Models ◽  
Unauthorized Access ◽  
Protection Method ◽  
Novel Method

In this paper, we propose a novel method for protecting convolutional neural network models with a secret key set so that unauthorized users without the correct key set cannot access trained models. The method enables us to protect not only from copyright infringement but also the functionality of a model from unauthorized access without any noticeable overhead. We introduce three block-wise transformations with a secret key set to generate learnable transformed images: pixel shuffling, negative/positive transformation, and format-preserving Feistel-based encryption. Protected models are trained by using transformed images. The results of experiments with the CIFAR and ImageNet datasets show that the performance of a protected model was close to that of non-protected models when the key set was correct, while the accuracy severely dropped when an incorrect key set was given. The protected model was also demonstrated to be robust against various attacks. Compared with the state-of-the-art model protection with passports, the proposed method does not have any additional layers in the network, and therefore, there is no overhead during training and inference processes.

scholarly journals PSAP: Improving Accuracy of Students' Final Grade Prediction using ID3 and C4.5

2019 ◽  
Vol 6 (2) ◽  
pp. 125-133
Author(s):  
Ismail Yusuf Panessai ◽  
Muhammad Modi Lakulu ◽  
Mohd Hishamuddin Abdul Rahman ◽  
Noor Anida Zaria Mohd Noor ◽  
Nor Syazwani Mat Salleh ◽  
...  
Keyword(s):  
Student Performance ◽  
Web Application ◽  
Confusion Matrix ◽  
Classification Model ◽  
Web Based ◽  
Final Grade ◽  
Grade Prediction ◽  
Qualitative Approaches ◽  
Improving Accuracy ◽  
The Web

PSAP: Improving Accuracy of Students' Final Grade Prediction using ID3 and C4.5 This study was aimed to increase the performance of the Predicting Student Academic Performance (PSAP) system, and the outcome is to develop a web application that can be used to analyze student performance during present semester. Development of the web-based application was based on the evolutionary prototyping model. The study also analyses the accuracy of the classifier that is constructed for the prediction features in the web application. Qualitative approaches by user evaluation questionnaire were used for this study. A number of few personnel expert users which are lecturers from Universiti Pendidikan Sultan Idris were chosen as respondents. Each respondent is instructed to answer a total of 27 questions regarding respondent’s background and web application design. The accuracy of the classifier for the prediction features is tested by using the confusion matrix by using the test set of 24 rows. The findings showed the views of respondents on the aspects of interface design, functionality, navigation, and reliability of the web-based application that is developed. The result also showed that accuracy for the classifier constructed by using ID3 classification model (C4.5) is 79.18% and the highest compared to Naïve Bayes and Generalized Linear classification model.

Two-Stage Adaptive Classification Cloud Workload Prediction Based on Neural Networks

2019 ◽  
Vol 11 (2) ◽  
pp. 1-23 ◽  
Author(s):  
Lei Li ◽  
Yilin Wang ◽  
Lianwen Jin ◽  
Xin Zhang ◽  
Huiping Qin
Keyword(s):  
Neural Networks ◽  
Prediction Models ◽  
Network Models ◽  
Classification Model ◽  
Neural Network Models ◽  
Two Stage ◽  
Workload Prediction ◽  
Random Mutation ◽  
Prediction Result ◽  
Adaptive Classification

Workload prediction is important for automatic scaling of resource management, and a high accuracy of workload prediction can reduce the cost and improve the resource utilization in the cloud. But, the task request is usually random mutation, so it is difficult to achieve more accurate prediction result for single models. Thus, to improve the prediction result, the authors proposed a novel two-stage workload prediction model based on artificial neural networks (ANNs), which is composed of one classification model and two prediction models. On the basis of the first-order gradient feature, the model can categorize the workload into two classes adaptively. Then, it can predict the workload by using the corresponding prediction neural network models according to the classification results. The experiment results demonstrate that the suggested model can achieve more accurate workload prediction compared with other models.

scholarly journals Design of MRI structured spiking neural networks and learning algorithms for personalized modelling, analysis, and prediction of EEG signals

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Samaneh Alsadat Saeedinia ◽  
Mohammad Reza Jahed-Motlagh ◽  
Abbas Tafakhori ◽  
Nikola Kasabov
Keyword(s):  
Information Integration ◽  
Learning Algorithm ◽  
Network Models ◽  
Eeg Signals ◽  
Neural Network Models ◽  
Modeling Analysis ◽  
Modelling Analysis ◽  
Eeg Data ◽  
Novel Method ◽  
Spike Time Dependent Plasticity

AbstractThis paper proposes a novel method and algorithms for the design of MRI structured personalized 3D spiking neural network models (MRI-SNN) for a better analysis, modeling, and prediction of EEG signals. It proposes a novel gradient-descent learning algorithm integrated with a spike-time-dependent-plasticity algorithm. The models capture informative personal patterns of interaction between EEG channels, contrary to single EEG signal modeling methods or to spike-based approaches which do not use personal MRI data to pre-structure a model. The proposed models can not only learn and model accurately measured EEG data, but they can also predict signals at 3D model locations that correspond to non-monitored brain areas, e.g. other EEG channels, from where data has not been collected. This is the first study in this respect. As an illustration of the method, personalized MRI-SNN models are created and tested on EEG data from two subjects. The models result in better prediction accuracy and a better understanding of the personalized EEG signals than traditional methods due to the MRI and EEG information integration. The models are interpretable and facilitate a better understanding of related brain processes. This approach can be applied for personalized modeling, analysis, and prediction of EEG signals across brain studies such as the study and prediction of epilepsy, peri-perceptual brain activities, brain-computer interfaces, and others.

scholarly journals Identification of toxic mold species through Raman spectroscopy of fungal conidia

PLoS ONE ◽  
2020 ◽  
Vol 15 (11) ◽  
pp. e0242361
Author(s):  
Benjamin D. Strycker ◽  
Zehua Han ◽  
Zheng Duan ◽  
Blake Commer ◽  
Kai Wang ◽  
...  
Keyword(s):  
Spectral Data ◽  
Raman Spectra ◽  
Network Models ◽  
Excitation Wavelength ◽  
Raman Signal ◽  
Neural Network Models ◽  
Species Classification ◽  
Melanin Biosynthesis ◽  
Raman Spectral Data ◽  
Raman Spectral

We use a 785 nm shifted excitation Raman difference (SERDS) technique to measure the Raman spectra of the conidia of 10 mold species of especial toxicological, medical, and industrial importance, including Stachybotrys chartarum, Penicillium chrysogenum, Aspergillus fumigatus, Aspergillus flavus, Aspergillus oryzae, Aspergillus niger, and others. We find that both the pure Raman and fluorescence signals support the hypothesis that for an excitation wavelength of 785 nm the Raman signal originates from the melanin pigments bound within the cell wall of the conidium. In addition, the major features of the pure Raman spectra group into profiles that we hypothesize may be due to differences in the complex melanin biosynthesis pathways. We then combine the Raman spectral data with neural network models to predict species classification with an accuracy above 99%. Finally, the Raman spectral data of all species investigated is made freely available for download and use.

Sign in / Sign up

Export Citation Format

Share Document