A mobile-phone-based diagnostic tool, which most of the population can easily access, could be a game changer in increasing the number of examinations of people with dental caries. This study aimed to apply a deep learning algorithm in diagnosing the stages of smooth surface caries via smartphone images. Materials and methods: A training dataset consisting of 1902 photos of the smooth surface of teeth taken with an iPhone 7 from 695 people was used. Four deep learning models, consisting of Faster Region-Based Convolutional Neural Networks (Faster R-CNNs), You Only Look Once version 3 (YOLOv3), RetinaNet, and Single-Shot Multi-Box Detector (SSD), were tested to detect initial caries lesions and cavities. The reference standard was the diagnosis of a dentist based on image examination according to the International Caries Classification and Management System (ICCMS) classification. Results: For cavitated caries, YOLOv3 and Faster R-CNN showed the highest sensitivity among the four tested models, at 87.4% and 71.4%, respectively. The sensitivity levels of these two models were only 36.9 % and 26% for visually non-cavitated (VNC). The specificity of the four models reached above 86% for cavitated caries and above 71% for VNC. Conclusion: The clinical application of YOLOv3 and Faster R-CNN models for diagnosing dental caries via smartphone images was promising. The current study provides a preliminary insight into the potential translation of AI from the laboratory to clinical practice.

Artificial neural network (ANN) has been well applied in pattern recognition, classification and machine learning thanks to its high performance. Most ANNs are designed by a static structure whose weights are trained during a learning process by supervised or unsupervised methods. These training methods require a set of initial weights values, which are normally randomly generated, with different initial sets of weight values leading to different convergent ANNs for the same training set. Dealing with these drawbacks, a trend of dynamic ANN was invoked in the past year. However, they are either too complex or far from practical applications such as in the pathology predictor in binary multi-input multi-output (MIMO) problems, when the role of a symptom is considered as an agent, a pathology predictor’s outcome is formed by action of active agents while other agents’ activities seem to be ignored or have mirror effects. In this paper, we propose a new dynamic structural ANN for MIMO problems based on the dependency graph, which gives clear cause and result relationships between inputs and outputs. The new ANN has the dynamic structure of hidden layer as a directed graph showing the relation between input, hidden and output nodes. The properties of the new dynamic structural ANN are experienced with a pathology problem and its learning methods’ performances are compared on a real well known dataset. The result shows that both approaches for structural learning process improve the quality of ANNs during learning iteration.

Dynamic multi-criteria decision-making (DMCDM) models have many meaningful applications in real life in which solving indeterminacy of information in DMCDMs strengthens the potential application of DMCDM. This study introduces an extension of dynamic internal-valued neutrosophic sets namely generalized dynamic internal-valued neutrosophic sets. Based on this extension, we develop some operators and a TOPSIS method to deal with the change of both criteria, alternatives, and decision-makers by time. In addition, this study also applies the proposal model to a real application that facilitates ranking students according to attitude-skill-knowledge evaluation model. This application not only illustrates the correctness of the proposed model but also introduces its high potential appliance in the education domain.

In dental diagnosis, recognizing tooth complications quickly from radiology (e.g., X-rays) takes highly experienced medical professionals. By using object detection models and algorithms, this work is much easier and needs less experienced medical practitioners to clear their doubts while diagnosing a medical case. In this paper, we propose a dental defect recognition model by the integration of Adaptive Convolution Neural Network and Bag of Visual Word (BoVW). In this model, BoVW is used to save the features extracted from images. After that, a designed Convolutional Neural Network (CNN) model is used to make quality prediction. To evaluate the proposed model, we collected a dataset of radiography images of 447 patients in Hanoi Medical Hospital, Vietnam, with third molar complications. The results of the model suggest accuracy of 84% ± 4%. This accuracy is comparable to that of experienced dentists and radiologists.

Studies on automatically predicting student learning outcomes often focus on developing and optimizing machine learning algorithms that fit the data captured from different education systems. This approach has a fatal weakness when it is used for disadvantaged groups, such as those with academic warnings or who have dropped out, because these groups are often much smaller than other common groups in number. The imbalanced data that have class distribution skew create a big challenge to training good classification models. The significant approach to tackle this challenge is applying oversampling methods to increase the number of minor classes; however, generating good new samples from the existing instances of a minor class is still a hard issue and requires new investigation. This study presents two new methods of handling data imbalance based on the original algorithms SMOTE and adaptive synthetic sampling (ADASYN), called Improved SMOTE (I_SMOTE) and Improved ADASYN (I_ADASYN). These modifications involve a new selecting fit candidate method based on a new similarity measurement and a roulette wheel selection to generate synthetic data samples. The aim is to rebalance data and therefore improve the prediction accuracy of minor groups. The proposal methods were designed and applied to education datasets, and they were tested on public datasets and a dataset collected from a Vietnamese university for evaluation. The experimental results on learning datasets showed the high potential of novel algorithms, I_SMOTE and I_ADASYN, for student academic performance problems in general and at‐risk student groups especially. Empirical results proved that the recall, precision, and F1‐score of the minority class of I_SMOTE and I_ADASYN are strongly better than the original balancing algorithms. Besides, the I_SMOTE and I_ADASYN also improve relatively by 6.6% and 8.0% of the ROC area compared to the original SMOTE and ADASYN, respectively.