A healthy life is essential for a happy society, however it is a fact that seemingly invisible diseases plague our families and people suffer. The thyroid disease falls in such a category. Thyroid ...disorders are long-term and with carefully handled illnesses, people with thyroid disorders may also live stable and normal lives. Thyroid diagnosis, particularly for an inexperienced clinician, is a difficult proposal. Many researchers have established various methods for the diagnosis of the disease and several models for disease prediction have been developed. As with several other domains, machine learning approaches to modelling health care problems is gaining popularity. This study aims at providing solutions towards such a thyroid disease prediction. Dimension reduction techniques are applied, and reduced dimension data input to classifiers. Also, data augmentation is applied so as to be able to generate sufficient data for deep neural network model. Classifier prediction is compared to other similar researches. Real life dataset for thyroid disease has been used, and experiments conducted in distributed environment. Our proposed two stage approach gives a maximum accuracy of 99.95% which is very good as compared to existing techniques. We have shown that dimension reduction and data augmentation can be used very efficiently for achieving high accuracy of disease prediction.
In the field of Brain Computer Interface (BCI), applications in real life like emotion recognition from recorded electrical activity from brain have become famous topic of research nowadays. Learning ...successful representations of consistent performances from electroencephalogram (EEG) signals is one of the difficulties in recognition tasks. This research is intended to propose a discriminative and efficacious classification approach for categorizing brain signals patterns depending on the level of activity or frequency for recognizing emotion states. The paper classifies three possible emotion states such as neutral, negative and positive emotional states by operating the Muse EEG headset with four electrode channels (AF7, AF8, TP9, TP10) captured while a subject was watching an emotional video clip on screen. In this experiment various statistical, linear and non linear features are extracted and then Machine and Deep learning based models are implemented to classify the EEG evoked emotions. In this work, a brief comparison study is carried out between the various implemented models with respect to train and test accuracy, recall, precision and F1 score. The highest average accuracy achieved are 98.13% for the proposed Convolutional Neural Network (CNN) model among all implemented Deep learning models and 98.12% for Random forest among the various machine learning techniques implemented. The proposed Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) model with 97.42 and 97.19% and Decision tree and Support Vector Machine with 96.25 and 96.42% have also provided comparable results for emotion classification respectively.
COVID-19 is an epidemic disease that has threatened all the people at worldwide scale and eventually became a pandemic It is a crucial task to differentiate COVID-19-affected patients from healthy ...patient populations. The need for technology enabled solutions is pertinent and this paper proposes a deep learning model for detection of COVID-19 using Chest X-Ray (CXR) images. In this research work, we provide insights on how to build robust deep learning based models for COVID-19 CXR image classification from Normal and Pneumonia affected CXR images. We contribute a methodical escort on preparation of data to produce a robust deep learning model. The paper prepared datasets by refactoring, using images from several datasets for ameliorate training of deep model. These recently published datasets enable us to build our own model and compare by using pre-trained models. The proposed experiments show the ability to work effectively to classify COVID-19 patients utilizing CXR. The empirical work, which uses a 3 convolutional layer based Deep Neural Network called “DeepCOVNet” to classify CXR images into 3 classes: COVID-19, Normal and Pneumonia cases, yielded an accuracy of 96.77% and a F1-score of 0.96 on two different combination of datasets.
The novel coronavirus illness (COVID-19) outbreak, which began in a seafood market in Wuhan, Hubei Province, China, in mid-December 2019, has spread to almost all countries, territories, and places ...throughout the world. And since the fault in diagnosis of a disease causes a psychological impact, this was very much visible in the spread of COVID-19. This research aims to address this issue by providing a better solution for diagnosis of the COVID-19 disease. The paper also addresses a very important issue of having less data for disease prediction models by elaborating on data handling techniques. Thus, special focus has been given on data processing and handling, with an aim to develop an improved machine learning model for diagnosis of COVID-19. Random Forest (RF), Decision tree (DT), K-Nearest Neighbor (KNN), Logistic Regression (LR), Support vector machine, and Deep Neural network (DNN) models are developed using the Hospital Israelita Albert Einstein (in São Paulo, Brazil) dataset to diagnose COVID-19. The dataset is pre-processed and distributed DT is applied to rank the features. Data augmentation has been applied to generate datasets for improving classification accuracy. The DNN model dominates overall techniques giving the highest accuracy of 96.99%, recall of 96.98%, and precision of 96.94%, which is better than or comparable to other research work. All the algorithms are implemented in a distributed environment on the Spark platform.
Brain magnetic resonance images (MRI) convey vital information for making diagnostic decisions and are widely used to detect brain tumors. This research proposes a self-supervised pre-training method ...based on feature representation learning through contrastive loss applied to unlabeled data. Self-supervised learning aims to understand vital features using the raw input, which is helpful since labeled data is scarce and expensive. For the contrastive loss-based pre-training, data augmentation is applied to the dataset, and positive and negative instance pairs are fed into a deep learning model for feature learning. Subsequently, the features are passed through a neural network model to maximize similarity and contrastive learning of the instances. This pre-trained model serves as an encoder for supervised training and then the classification of MRI images. Our results show that self-supervised pre-training with contrastive loss performs better than random or ImageNet initialization. We also show that contrastive learning performs better when the diversity of images in the pre-training dataset is more. We have taken three differently sized ResNet models as the base models. Further, experiments were also conducted to study the effect of changing the augmentation types for generating positive and negative samples for self-supervised training.
We perceive big data with massive datasets of complex and variegated structures in the modern era. Such attributes formulate hindrances while analyzing and storing the data to generate apt ...aftermaths. Privacy and security are the colossal perturb in the domain space of extensive data analysis. In this paper, our foremost priority is the computing technologies that focus on big data, IoT (Internet of Things), Cloud Computing, Blockchain, and fog computing. Among these, Cloud Computing follows the role of providing on-demand services to their customers by optimizing the cost factor. AWS, Azure, Google Cloud are the major cloud providers today. Fog computing offers new insights into the extension of cloud computing systems by procuring services to the edges of the network. In collaboration with multiple technologies, the Internet of Things takes this into effect, which solves the labyrinth of dealing with advanced services considering its significance in varied application domains. The Blockchain is a dataset that entertains many applications ranging from the fields of crypto-currency to smart contracts. The prospect of this research paper is to present the critical analysis and review it under the umbrella of existing extensive data systems. In this paper, we attend to critics' reviews and address the existing threats to the security of extensive data systems. Moreover, we scrutinize the security attacks on computing systems based upon Cloud, Blockchain, IoT, and fog. This paper lucidly illustrates the different threat behaviour and their impacts on complementary computational technologies. The authors have mooted a precise analysis of cloud-based technologies and discussed their defense mechanism and the security issues of mobile healthcare.
The software engineering community is working to develop reliable metrics to improve software quality. It is estimated that understanding the source code accounts for 60% of the software maintenance ...effort. Cognitive informatics is important in quantifying the degree of difficulty or the efforts made by developers to understand the source code. Several empirical studies were conducted in 2003 to assign cognitive weights to each possible basic control structure of software, and these cognitive weights are used by several researchers to evaluate the cognitive complexity of software systems. In this paper, an effort has been made to categorize the Control Flow Graphs (CFGs) nodes according to their node features. In our case, we extracted seven unique features from the program, and each unique feature was assigned an integer value that we evaluated through Cognitive Complexity Measures (CCMs). We then incorporated CCMs' results as a node feature value in CFGs and generated the same based on the node connectivity for a graph. In order to obtain the feature representation of the graph, a node vector matrix is then created for the graph and passed to the Graph Convolutional Network (GCN). We prepared our data sets using GCN output and then built Deep Neural Network Defect Prediction (DNN-DP) and Convolutional Neural Network Defect Prediction (CNN-DP) models to predict software defects. The Python programming language is used, along with Keras and TensorFlow. Three hundred twenty Python programs were written by our talented UG and PG students, and all experiments were carried out during laboratory classes. Together with three skilled lab programmers, they compiled and ran each individual program and detected defect/no-defect programs before categorizing them into three different classes, namely Simple, Medium, and Complex programs. Accuracy, Receiver Operating Characteristics (ROC), Area Under Curve (AUC), F-measure, Precision and hyper-parameter tuning procedures are used to evaluate the approaches. The experimental results show that the proposed models outperformed state-of-the-art methods such as Nave Bayes (NB), Decision Tree (DT), Support Vector Machine (SVM), and Random Forest (RF) in all evaluation criteria.
StockNet—GRU based stock index prediction Gupta, Umang; Bhattacharjee, Vandana; Bishnu, Partha Sarathi
Expert systems with applications,
11/2022, Volume:
207
Journal Article
Peer reviewed
Predicting financial trends of stock indexes is important for investors to reduce risk on investment and efficient decision making if the prediction is made accurately. Researchers, in recent times ...have applied deep learning approaches in this field which have essentially beaten conventional machine learning approaches. To overcome the issue of overfitting we presented a new data augmentation approach in our GRU based StockNet model consisting of two modules. Injection module to prohibit overfitting and Investigation module for stock index forecasting. The proposed approach has been validated on Indian stock market (CNX-Nifty). Proposed StockNet-c model produces 65.59%, 27.30% and 14.89 % less test loss in terms of RMSE, MAE and MAPE respectively, in comparison to TargetNet model where overfitting prohibition injection module is missing.
•Stock index prediction with data augmentation approach to prohibit overfitting.•StockNet model with GRU network with injection and investigation modules.•Proposed models shows the best result in CNX Nifty dataset.•The prediction results of the proposed model have statistical significance.
•A Capsule Network (EEGCapsNet) is proposed to classify visually evoked EEG signals.•EEG signal is converted to spectrogram images using STFT transformation.•Overcomes the drawbacks of CNN ...architecture mainly lack of retaining spatial features and overfitting.•Experimentation is done on two datasets: Perceive and MindBig.•Implements 8 variations of Capsule Network models with overall classification accuracy, precision, recall, and F1 score.
Automated visual stimuli evoked multi-channel electroencephalograph (EEG) signals classification is in a nascent stage but is receiving progressive attention from researchers. The conventional techniques existing for EEG classification tasks overlook the spatial attributes of EEG signals, which contain spatial data information to characterize an image from visual stimuli evoked EEG signal. In this paper, a Deep learning structure implemented on STFT (Short Term Fourier Transform) generated spectrogram images and a Capsule Network (EEGCapsNet) is proposed. In this architecture, the time and frequency domain as well as, spatial attributes of the multi-channel EEG signals are extracted to build the spectrogram image and are fed to the proposed EEGCapsNet for classifying those EEG signals which are acquired from the stimuli evoked visual experience while seeing an image. For this purpose, two different EEG datasets (namely Perceive and MindBig) are trained on the proposed network. The highest average accuracy of 81.59% and 84.62% is reported for the proposed EEGCapsNet.
In recent studies, machine learning and deep learning strategies have been explored in many EEG-based application for best performance. More specifically, convolutional neural networks (CNNs) have ...demonstrated incredible capacity in electroencephalograph (EEG)-evoked emotion classification tasks. In preexisting case, CNN-based emotion classification techniques using EEG signals mostly involve a moderately intricate phase of feature extrication before any network model implementation. The CNNs are not able to well describe the natural interrelation among the various EEG channels, which basically provides essential data for the classification of different emotion states. In this paper, an efficacious and advanced version of CNN called Emotion-based Capsule Network (EmotionCapsNet) for multi-channel EEG-based emotion classification to achieve better classification accuracy is presented. EmotionCapsNet has been applied to the raw EEG signals as well as 2D image representation generated from EEG signals which can extricate descriptive and complex features from the EEG signals and decide the different emotional states. The proposed system is then compared with the other conventional machine learning and deep learning-based CNN model. Our strategy accomplishes an average accuracy of 77.50%, 78.44% and 79.38% for valence, arousal and dominance on the DEAP, 79.06%, 78.90% and 79.69% on AMIGOS and attains an average accuracy of 80.34%, 83.04% and 82.50% for valence, arousal and dominance on the DREAMER, respectively. These outcomes demonstrate that adapted strategy yields comparable precision on raw EEG signal and it also provides better classification results on spatiotemporal feature of EEG signal for emotion classification task.
PDF
