User Consumption Behavior Recognition Based on SMOTE and Improved AdaBoost

The sudden outbreak of COVID-19 has dealt a huge blow to traditional education and training companies. Institutions use the WeChat platform to attract users, but how to identify high-quality users has always been a difficult point for enterprises. In this paper, researchers proposed a classification algorithm based on SMOTE and the improved AdaBoost, which fuses feature information weights and sample weights to effectively solve the problems of overfitting and sample imbalance. To justify the study, it was compared with other traditional machine-learning algorithms. The accuracy and recall of the model increased by 19% and 36%, respectively, and the AUC value reached 0.98, indicating that the model could effectively identify the user's purchase intention. The proposed algorithm also ensures that it works well in spam identification and fraud detection. This research is of great significance for educational institutions to identify high-quality users of the WeChat platform and increase purchase conversion rate.

Medical Diagnosis and Identification of Covid -19 by Intelligent IoT System and Resnet 18 Bilinear Deep Greedy Network

An international health crisis has been caused by the widespread COVID-19 epidemic. COVID-19 patient diagnoses are made using deep learning, although this necessitates a massive radiography data collection in order to efficiently deliver an optimum result. This paper presents a novel Intelligent System with IoT sensors for covid 19 and "Bilinear Resnet 18 Deep Greedy Network,"which is effective with a limited amount of datasets. Despite peculiarities brought on by a small dataset, the suggested approach could successfully combat the anomalies of over fitting and under fitting. The suggested architecture ensures a successful conclusion when the trained model is correctly evaluated using the provided X-ray datasets of COVID-19 cases. The recommended model offers accuracy of 97%, which is superior to existing methodologies. Better precision, recall, and F1 score are provided;which are 98%, 96%, and 96.94% respectively, which is better than other existing methodology. © 2022 IEEE.

Information Systems in Medical Settings: A Covid-19 Detection System Using X-Ray Scans

Beginning in 2020, the new coronavirus began to expand globally. Due to Covid-19, millions of individuals are infected. Initially, the availability of corona test kits was problematic. Researchers examined the present scenario and developed the Covid-19 X-ray scan detection system. In terms of Covid-19 detection, artificial intelligence (AI)-based solutions give superior outcomes. Many AI-based models can not provide optimum results because of the issue of overfitting, which has a direct impact on model efficiency. In this work, we developed the CNN-based classification method based on the pre-trained Inception-v3 for normal, viral pneumonia, lung opacity, and Covid-19 samples. In the suggested model, we employed transfer learning to produce promising results for binary class classification. The presented model attained impressive outcomes with an accuracy of 99.42% for Covid-19 vs. Normal, 99.01% for Covid-19 vs. Lung Opacity, and 99.8% for Covid-19 vs. Viral Pneumonia, and 99.93% for Lung Opacity vs. Viral Pneumonia. Comparing the suggested model to existing deep learning-based systems indicated that ours was better. © 2022 IEEE.

FADCOVNET: Fast Automatic Detection Covid-19 Based on Inception-Resnet-V2 Model

Covid-19 is still a threat to human health. Initial handling in detecting the status of positive COVID-19 patients or not through the IT sector is still very much needed to help the government control the covid-19 outbreak. This study offers a new framework of deep learning classification to help radiologists work in auto-detecting cases of COVID-19 by processing patient X-Ray chest (we call it FADCOVNET). By combining pre-processing techniques with a modified Inception Resnet V2 trained network on the Fully Connected layer and by adding pre-processing data. To control overfitting, the data augmentation method is used. The FADCOVNET model will be compared with the transfer learning model (Resnet50, Inception V3, Inception-Resnet-V3).The dataset used in this study is chest X-ray data for COVID cases as many as 4369 total data. In addition, this study also tested the performance of FADCOVNET on the Covid and healthy chest CT-Scan dataset of 8467 total data. The test results show that the performance of FADCOVNET on the accuracy, sensitivity, specification, precision, and F1-Score are 97%, 98%, 97%, 95%, and 96%, respectively. The results obtained outperform other transfer models. while the accuracy obtained from testing with the CT Scan dataset is 97%. This proves that the FADCOVNET model that we have built can ensure the generalizability of the model very well. From this test, it can be concluded that the proposed CNN architecture works very well in detecting COVID-19. © 2022 IEEE.

Using Ultrasound Image Augmentation and Ensemble Predictions to Prevent Machine-Learning Model Overfitting.

Deep learning predictive models have the potential to simplify and automate medical imaging diagnostics by lowering the skill threshold for image interpretation. However, this requires predictive models that are generalized to handle subject variability as seen clinically. Here, we highlight methods to improve test accuracy of an image classifier model for shrapnel identification using tissue phantom image sets. Using a previously developed image classifier neural network-termed ShrapML-blind test accuracy was less than 70% and was variable depending on the training/test data setup, as determined by a leave one subject out (LOSO) holdout methodology. Introduction of affine transformations for image augmentation or MixUp methodologies to generate additional training sets improved model performance and overall accuracy improved to 75%. Further improvements were made by aggregating predictions across five LOSO holdouts. This was done by bagging confidences or predictions from all LOSOs or the top-3 LOSO confidence models for each image prediction. Top-3 LOSO confidence bagging performed best, with test accuracy improved to greater than 85% accuracy for two different blind tissue phantoms. This was confirmed by gradient-weighted class activation mapping to highlight that the image classifier was tracking shrapnel in the image sets. Overall, data augmentation and ensemble prediction approaches were suitable for creating more generalized predictive models for ultrasound image analysis, a critical step for real-time diagnostic deployment.

Pneumonia image classification based on convolutional neural network

This letter introduces an improved convolutional neural network (CNN), which is used to classify and recognize different types of pneumonia using chest CT images. This classifying model is built and trained on thousands of real clinical chest CT images, which respectively belong to patients with viral pneumonia, patients with bacterial pneumonia, patients with COVID-19, and nonpatients. To richen the dataset and avoid over-fitting, pre-processing methods are recommended. Then the paper elaborates the structure of the new network and compares the performance of different optimizers in this dataset. Finally, the accuracy, specificity, precision, sensitivity, and F1-score of the model are calculated to quantitatively evaluate the performance of this model. The final training accuracy is about 97.9%, and the test accuracy is 91.8%. © 2022 IEEE.

PGUNET: COVID-19 CT IMAGE SEGMENTATION USING GAN AND FEATURE PYRAMID

Since computed tomography (CT) provides the most sensitive radiological technique for diagnosing COVID-19, CT has been used as an efficient and necessary aided diagnosis. However, the size and number of publicly available COVID-19 imaging datasets are limited and have problems such as low data volume, easy overfitting for training, and significant differences in the characteristics of lesions at different scales. Our work presents an image segmentation network, Pyramid-and-GAN-UNet (PGUNet), to support the segmentation of COVID-19 lesions by combining feature pyramid and generative adversarial network (GAN). Using GAN, the segmentation network can learn more abundant high-level features and increase the generalization ability. The module of the feature pyramid is used to solve the differences between image features at different levels. Compared with the current mainstream method, our experimental results show that the proposed network achieved more competitive performances on the CT slice datasets of the COVID-19 CT Segmentation dataset and CC-CCII dataset. © 2022 IEEE.

Face Authentication from Masked Face Images Using Deep Learning on Periocular Biometrics

Nowadays, identity theft is an alarming issue with the growth of e-commerce and online services. Moreover, due to the Covid-19 pandemic, society has been pushed towards the usage of masks for people to safely interact with one another. It is hard to recognize a person if the face is mostly covered, even more so to artificial intelligence who have more difficulty identifying a masked individual. To further protect personal information and to develop a secure information system, more comprehensive bio-metric approaches are required. The currently used facial recognition systems are using biometrics such as periocular regions, iris, face, skin tone and racial information etc. In this paper, we apply a deep learning-based authentication approach using periocular biometric information to enhance the performance of the facial recognition system. We used the Real-World Masked Face Dataset (RMFD) and other datasets to develop our system. We implemented some experiments using CNN model on the periocular region information of the images. Hence, we developed a system that can recognize a person from only using a small region of face, which in this case is the periocular information including both eyes and eyebrows region. There is only a focus on the periocular region with our model in the view of the fact that the periocular region of the face is the main reliable source of information we can get while a person is wearing a face mask. © 2022, Springer Nature Switzerland AG.

The Model of Regularization Coefficient in Polynomial Regression for Modelling the Spread of COVID-19 in Romania

Research on the pandemic situation of COVID-19 is very important for delivering detailed risk analyzes based on estimating the peak of the pandemic. The machine learning approach has a major role to play in predicting the number of COVID-19 cases. Most research on COVID-19 uses polynomial regression for analysis. When a regression model is build, often, the model fails to generalize on unseen data. For instance, the model might end up becoming too complex, having significantly high variance due to over-fitting, thereby impacting the model performance on new data sets. To avoid over-fitting of the polynomial regression, a regularization method can be used to suppress the coefficients of the higher order polynomial, a principle that allows the smoothness of the regression function. The aim of this paper is to formulate a mathematical model for regularization coefficient in polynomial regression and evaluate this approach to enable obtaining meaningful results on a COVID-19 data set. Therefore we believe that our results will contribute to a better understanding of the over-fitting process in polynomial regression. Our methodology consists of following major steps: i) optimizing the model using k-fold cross-validation for finding an optimal regularization coefficient and ii) comparing the performance of ridge regression and lasso regression using accuracy metrics. Moreover, our approach could also have a potential impact in machine learning education, regarding the understanding of the underlying mathematical machinery behind polynomial regression algorithms. The obtained results show that the polynomial model built using lasso regression, outperforms the ridge regression. © 2022 IEEE.

COVID-19 Image Diagnosis on CT Images Using Deep Learning

A dangerous and contagious respiratory disease, COVID-19, is currently spreading quickly around the world. One of the important approaches to slow down the spread is to detect those who have COVID-19 and ask them to self-quarantine. The hospitals use CT scans to diagnose COVID-19, and Artificial Intelligence could diagnose as well. Our study collects around 300 images including both those belong to COVID-19 patients and normal people from Kaggle. We run both CNN and FFNN models on the data and record the accuracy together with the F1 score of each model. It turns out that the accuracy for CNN is 100%, while for FFNN is 96.88%. CNN has a lower testing loss, and it takes less time to train. The result has shown that these models could accurately predict the correct one, but there are some drawbacks. The 100% accuracy might indicate overfitting. Things that the study could improve upon include collecting more data and adding new classifications. © 2021 IEEE.

On the Generation of Medical Dialogs for COVID-19

Under the pandemic of COVID-19, people experiencing COVID19-related symptoms have a pressing need to consult doctors. Because of the shortage of medical professionals, many people cannot receive online consultations timely. To address this problem, we aim to develop a medical dialog system that can provide COVID19-related consultations. We collected two dialog datasets - CovidDialog - (in English and Chinese respectively) containing conversations between doctors and patients about COVID-19. While the largest of their kind, these two datasets are still relatively small compared with generaldomain dialog datasets. Training complex dialog generation models on small datasets bears high risk of overfitting. To alleviate overfitting, we develop a multi-task learning approach, which regularizes the data-deficient dialog generation task with a masked token prediction task. Experiments on the CovidDialog datasets demonstrate the effectiveness of our approach. We perform both human evaluation and automatic evaluation of dialogs generated by our method. Results show that the generated responses are promising in being doctorlike, relevant to conversation history, clinically informative and correct. © 2021 Association for Computational Linguistics.

Optimal hyperparameter selection of deep learning models for COVID-19 chest X-ray classification.

The first and most critical response to curbing the spread of the novel coronavirus disease (COVID-19) is to deploy effective techniques to test potentially infected patients, isolate them and commence immediate treatment. However, several test kits currently in use are slow and in a shortage of supply. This paper presents techniques for diagnosing COVID-19 from chest X-ray (CXR) and address problems associated with training deep models with less voluminous datasets and class imbalance as obtained in most available CXR datasets on COVID-19. We used the discriminative fine-tuning approach, which dynamically assigns different learning rates to each layer of the network. The learning rate is set using the cyclical learning rate policy that changes per iteration. This flexibility ensured rapid convergence and avoided being stuck in saddle point plateau. In addition, we addressed the high computational demand of deep models by implementing our algorithm using the memory- and computational-efficient mixed-precision training. Despite the availability of scanty datasets, our model achieved high performance and generalisation. A Validation accuracy of 96.83%, sensitivity and specificity of 96.26% and 95.54% were obtained, respectively. When tested on an entirely new dataset, the model achieves 97% accuracy without further training. Lastly, we presented a visual interpretation of the model's output to prove that the model can aid radiologists in rapidly screening for the symptoms of COVID-19.