Your search

Text classification is an important topic in natural language processing, with the development of social network, many question-and-answer pairs regarding health-care and medicine flood social platforms. It is of great social value to mine and classify medical text and provide targeted medical services for patients. The existing algorithms of text classification can deal with simple semantic text, especially in the field of Chinese medical text, the text structure is complex and includes a large number of medical nomenclature and professional terms, which are difficult for patients to understand. We propose a Chinese medical text classification model using a BERT-based Chinese text encoder by N-gram representations (ZEN) and capsule network, which represent feature uses the ZEN model and extract the features by capsule network, we also design a N-gram medical dictionary to enhance medical text representation and feature extraction. The experimental results show that the precision, recall and F1-score of our model are improved by 10.25%, 11.13% and 12.29%, respectively, compared with the baseline models in average, which proves that our model has better performance.

The number of tourist attractions reviews, travel notes and other texts has grown exponentially in the Internet age. Effectively mining users’ potential opinions and emotions on tourist attractions, and helping to provide users with better recommendation services, which is of great practical significance. This paper proposes a multi-channel neural network model called Pre-BiLSTM combined with a pre-training mechanism. The model uses a combination of coarse and fine- granularity strategies to extract the features of text information such as reviews and travel notes to improve the performance of text sentiment analysis. First, we construct three channels and use the improved BERT and skip-gram methods with negative sampling to vectorize the word-level and vocabulary-level text, respectively, so as to obtain more abundant textual information. Second, we use the pre-training mechanism of BERT to generate deep bidirectional language representation relationships. Third, the vectors of the three channels are input into the BiLSTM network in parallel to extract global and local features. Finally, the model fuses the text features of the three channels and classifies them using SoftMax classifier. Furthermore, numerical experiments are conducted to demonstrate that Pre-BiLSTM outperforms the baselines by 6.27%, 12.83% and 18.12% in average in terms of accuracy, precision and F1-score.

Recently, a lot of Chinese patients consult treatment plans through social networking platforms, but the Chinese medical text contains rich information, including a large number of medical nomenclatures and symptom descriptions. How to build an intelligence model to automatically classify the text information consulted by patients and recommend the correct department for patients is very important. In order to address the problem of insufficient feature extraction from Chinese medical text and low accuracy, this paper proposes a dual channel Chinese medical text classification model. The model extracts feature of Chinese medical text at different granularity, comprehensively and accurately obtains effective feature information, and finally recommends departments for patients according to text classification. One channel of the model focuses on medical nomenclatures, symptoms and other words related to hospital departments, gives different weights, calculates corresponding feature vectors with convolution kernels of different sizes, and then obtains local text representation. The other channel uses the BiGRU network and attention mechanism to obtain text representation, highlighting the important information of the whole sentence, that is, global text representation. Finally, the model uses full connection layer to combine the representation vectors of the two channels, and uses Softmax classifier for classification. The experimental results show that the accuracy, recall and F1-score of the model are improved by 10.65%, 8.94% and 11.62% respectively compared with the baseline models in average, which proves that our model has better performance and robustness.

In view of the complex marine environment of navigation, especially in the case of multiple static and dynamic obstacles, the traditional obstacle avoidance algorithms applied to unmanned surface vehicles (USV) are prone to fall into the trap of local optimization. Therefore, this paper proposes an improved artificial potential field (APF) algorithm, which uses 5G communication technology to communicate between the USV and the control center. The algorithm introduces the USV discrimination mechanism to avoid the USV falling into local optimization when the USV encounter different obstacles in different scenarios. Considering the various scenarios between the USV and other dynamic obstacles such as vessels in the process of performing tasks, the algorithm introduces the concept of dynamic artificial potential field. For the multiple obstacles encountered in the process of USV sailing, based on the International Regulations for Preventing Collisions at Sea (COLREGS), the USV determines whether the next step will fall into local optimization through the discrimination mechanism. The local potential field of the USV will dynamically adjust, and the reverse virtual gravitational potential field will be added to prevent it from falling into the local optimization and avoid collisions. The objective function and cost function are designed at the same time, so that the USV can smoothly switch between the global path and the local obstacle avoidance. The simulation results show that the improved APF algorithm proposed in this paper can successfully avoid various obstacles in the complex marine environment, and take navigation time and economic cost into account.

Convolutional neural network (CNN) model based on deep learning has excellent performance for target detection. However, the detection effect is poor when the object is circular or tubular because most of the existing object detection methods are based on the traditional rectangular box to detect and recognize objects. To solve the problem, we propose the circular representation structure and RepVGG module on the basis of CenterNet and expand the network prediction structure, thus proposing a high-precision and high-efficiency lightweight circular object detection method RebarDet. Specifically, circular tubular type objects will be optimized by replacing the traditional rectangular box with a circular box. Second, we improve the resolution of the network feature map and the upper limit of the number of objects detected in a single detect to achieve the expansion of the network prediction structure, optimized for the dense phenomenon that often occurs in circular tubular objects. Finally, the multibranch topology of RepVGG is introduced to sum the feature information extracted by different convolution modules, which improves the ability of the convolution module to extract information. We conducted extensive experiments on rebar datasets and used AB-Score as a new evaluation method to evaluate RebarDet. The experimental results show that RebarDet can achieve a detection accuracy of up to 0.8114 and a model inference speed of 6.9 fps while maintaining a moderate amount of parameters, which is superior to other mainstream object detection models and verifies the effectiveness of our proposed method. At the same time, RebarDet’s high precision detection of round tubular objects facilitates enterprise intelligent manufacturing processes.

Objective. As the preclinical stage of Alzheimer’s disease (AD), Mild Cognitive Impairment (MCI) is characterized by hidden onset, which is difficult to detect early. Traditional neuropsychological scales are main tools used for assessing MCI. However, due to its strong subjectivity and the influence of many factors such as subjects’ educational background, language and hearing ability, and time cost, its accuracy as the standard of early screening is low. Therefore, the purpose of this paper is to propose a new key technology of fast digital early warning for MCI based on eye movement objective data analysis. Methodology. Firstly, four exploratory indexes (test durations, correlation degree, lengths of gaze trajectory, and drift rate) of MCI early warning are determined based on the relevant literature research and semistructured expert interview; secondly, the eye movement state is captured based on the eye tracker to realize the data extraction of four exploratory indexes. On this basis, the human-computer interactive 2.5-minute fast digital early warning paradigm for MCI is designed; thirdly, the rationality of the four early warning indexes proposed in this paper and their early warning effectiveness on MCI are verified. Results. Through the small sample test of human-computer interactive 2.5 fast digital early warning paradigm for MCI conducted by 32 elderly people aged 70–90 in a medical institution in Hangzhou, the two indexes of “correlation degree” and “drift rate” with statistical differences are selected. The experiment results show that AUC of this MCI early warning paradigm is 0.824. Conclusion. The key technology of human-computer interactive 2.5 fast digital early warning for MCI proposed in this paper overcomes the limitations of the existing MCI early warning tools, such as low objectification level, high dependence on professional doctors, long test time, requiring high educational level, and so on. The experiment results show that the early warning technology, as a new generation of objective and effective digital early warning tool, can realize 2.5-minute fast and high-precision preliminary screening and early warning for MCI in the elderly.

With the fifth generation (5G) communication technology, the mobile multiuser networks have developed rapidly. In this paper, the performance analysis of mobile multiuser networks which utilize decode-and-forward (DF) relaying is considered. We derive novel outage probability (OP) expressions. To improve the OP performance, we study the power allocation optimization problem. To solve the optimization problem, we propose an intelligent power allocation optimization algorithm based on grey wolf optimization (GWO). We compare the proposed GWO approach with three existing algorithms. The experimental results reveal that the proposed GWO algorithm can achieve a smaller OP, thus improving system efficiency. Also, compared with other channel models, the OP values of the 2-Rayleigh model are increased by 81.2% and 66.6%, respectively.

Classification of electroencephalogram (EEG) is a key approach to measure the rhythmic oscillations of neural activity, which is one of the core technologies of brain-computer interface systems (BCIs). However, extraction of the features from non-linear and non-stationary EEG signals is still a challenging task in current algorithms. With the development of artificial intelligence, various advanced algorithms have been proposed for signal classification in recent years. Among them, deep neural networks (DNNs) have become the most attractive type of method due to their end-to-end structure and powerful ability of automatic feature extraction. However, it is difficult to collect large-scale datasets in practical applications of BCIs, which may lead to overfitting or weak generalizability of the classifier. To address these issues, a promising technique has been proposed to improve the performance of the decoding model based on data augmentation (DA). In this article, we investigate recent studies and development of various DA strategies for EEG classification based on DNNs. The review consists of three parts: what kind of paradigms of EEG-based on BCIs are used, what types of DA methods are adopted to improve the DNN models, and what kind of accuracy can be obtained. Our survey summarizes the current practices and performance outcomes that aim to promote or guide the deployment of DA to EEG classification in future research and development.

To solve the problem of one-sided pursuit of the shortest distance but ignoring the tourist experience in the process of tourism route planning, an improved ant colony optimization algorithm is proposed for tourism route planning. Contextual information of scenic spots significantly effect people’s choice of tourism destination, so the pheromone update strategy is combined with the contextual information such as weather and comfort degree of the scenic spot in the process of searching the global optimal route, so that the pheromone update tends to the path suitable for tourists. At the same time, in order to avoid falling into local optimization, the sub-path support degree is introduced. The experimental results show that the optimized tourism route has greatly improved the tourist experience, the route distance is shortened by 20.5% and the convergence speed is increased by 21.2% compared with the basic algorithm, which proves that the improved algorithm is notably effective.

There are a large number of symptom consultation texts in medical and healthcare Internet communities, and Chinese health segmentation is more complex, which leads to the low accuracy of the existing algorithms for medical text classification. The deep learning model has advantages in extracting abstract features of text effectively. However, for a large number of samples of complex text data, especially for words with ambiguous meanings in the field of Chinese medical diagnosis, the word-level neural network model is insufficient. Therefore, in order to solve the triage and precise treatment of patients, we present an improved Double Channel (DC) mechanism as a significant enhancement to Long Short-Term Memory (LSTM). In this DC mechanism, two channels are used to receive word-level and char-level embedding, respectively, at the same time. Hybrid attention is proposed to combine the current time output with the current time unit state and then using attention to calculate the weight. By calculating the probability distribution of each timestep input data weight, the weight score is obtained, and then weighted summation is performed. At last, the data input by each timestep is subjected to trade-off learning to improve the generalization ability of the model learning. Moreover, we conduct an extensive performance evaluation on two different datasets: cMedQA and Sentiment140. The experimental results show that the DC-LSTM model proposed in this paper has significantly superior accuracy and ROC compared with the basic CNN-LSTM model.