recurrent neural network
Recently Published Documents





2022 ◽  
Vol 40 (1) ◽  
pp. 1-27
Agnès Mustar ◽  
Sylvain Lamprier ◽  
Benjamin Piwowarski

When conducting a search task, users may find it difficult to articulate their need, even more so when the task is complex. To help them complete their search, search engine usually provide query suggestions. A good query suggestion system requires to model user behavior during the search session. In this article, we study multiple Transformer architectures applied to the query suggestion task and compare them with recurrent neural network (RNN)-based models. We experiment Transformer models with different tokenizers, with different Encoders (large pretrained models or fully trained ones), and with two kinds of architectures (flat or hierarchic). We study the performance and the behaviors of these various models, and observe that Transformer-based models outperform RNN-based ones. We show that while the hierarchical architectures exhibit very good performances for query suggestion, the flat models are more suitable for complex and long search tasks. Finally, we investigate the flat models behavior and demonstrate that they indeed learn to recover the hierarchy of a search session.

Machines ◽  
2022 ◽  
Vol 10 (1) ◽  
pp. 63
Xinyong Zhang ◽  
Liwei Sun ◽  
Lingtong Qi

The optical-mechanical system of a space camera is composed of several complex components, and the effects of several factors (weight, gravity, modal frequency, temperature, etc.) on its system performance need to be considered during ground tests, launch, and in-orbit operation. In order to meet the system specifications of the optical camera system, the dimensional parameters of the optical camera structure need to be optimized. There is a highly nonlinear functional relationship between the dimensional parameters of the optical machine structure and the design indexes. The traditional method takes a significant amount of time for finite element calculation and is less efficient. In order to improve the optimization efficiency, a recurrent neural network prediction model based on the Bayesian regularization algorithm is proposed in this paper, and the NSGA-II is used to globally optimize multiple prediction objectives of the prediction model. The reflector of the space camera is used as an example to predict the weight, first-order modal frequency, and gravitational mirror deformation root mean square of the reflector, and to complete the lightweight design. The results show that the prediction model established by BR-RNN-NSGA-II offers high prediction accuracy for the design indexes of the reflector, which all reach over 99.6%, and BR-RNN-NSGA-II can complete the multi-objective optimization search efficiently and accurately. This paper provides a new idea of optimization of optical machine structure, which enriches the theory of complex structure design.

2022 ◽  
Vol 12 (2) ◽  
pp. 749
Yunfei Gao ◽  
Albert No

Finding a biomarker that indicates the subject’s age is one of the most important topics in biology. Several recent studies tried to extract a biomarker from brain imaging data including fMRI data. However, most of them focused on MRI data, which do not provide dynamics and lack attempts to apply recently proposed deep learning models. We propose a deep neural network model that estimates the age of a subject from fMRI images using a recurrent neural network (RNN), more precisely, a gated recurrent unit (GRU). However, applying neural networks is not trivial due to the high dimensional nature of fMRI data. In this work, we propose a novel preprocessing technique using the Automated Anatomical Labeling (AAL) atlas, which significantly reduces the input dimension. The proposed dimension reduction technique allows us to train our model with 640 training and validation samples from different projects under mean squared error (MSE). Finally, we obtain the correlation value of 0.905 between the predicted age and the actual age on 155 test samples. The proposed model estimates the age within the range of ±12 on most of the test samples. Our model is written in Python and is freely available for download.

2022 ◽  
Zihang Lei ◽  
Mengxi Jiang ◽  
Guangsong Yang ◽  
Tianmin Guan ◽  
Peng Huang ◽  

Stylianos Paraschiakos ◽  
Cláudio Rebelo de Sá ◽  
Jeremiah Okai ◽  
P. Eline Slagboom ◽  
Marian Beekman ◽  

AbstractThrough the quantification of physical activity energy expenditure (PAEE), health care monitoring has the potential to stimulate vital and healthy ageing, inducing behavioural changes in older people and linking these to personal health gains. To be able to measure PAEE in a health care perspective, methods from wearable accelerometers have been developed, however, mainly targeted towards younger people. Since elderly subjects differ in energy requirements and range of physical activities, the current models may not be suitable for estimating PAEE among the elderly. Furthermore, currently available methods seem to be either simple but non-generalizable or require elaborate (manual) feature construction steps. Because past activities influence present PAEE, we propose a modeling approach known for its ability to model sequential data, the recurrent neural network (RNN). To train the RNN for an elderly population, we used the growing old together validation (GOTOV) dataset with 34 healthy participants of 60 years and older (mean 65 years old), performing 16 different activities. We used accelerometers placed on wrist and ankle, and measurements of energy counts by means of indirect calorimetry. After optimization, we propose an architecture consisting of an RNN with 3 GRU layers and a feedforward network combining both accelerometer and participant-level data. Our efforts included switching mean to standard deviation for down-sampling the input data and combining temporal and static data (person-specific details such as age, weight, BMI). The resulting architecture produces accurate PAEE estimations while decreasing training input and time by a factor of 10. Subsequently, compared to the state-of-the-art, it is capable to integrate longer activity data which lead to more accurate estimations of low intensity activities EE. It can thus be employed to investigate associations of PAEE with vitality parameters of older people related to metabolic and cognitive health and mental well-being.

2022 ◽  
Vol 12 (1) ◽  
pp. 504
Abdul Razaque ◽  
Bandar Alotaibi ◽  
Munif Alotaibi ◽  
Shujaat Hussain ◽  
Aziz Alotaibi ◽  

People who use social networks often fall prey to clickbait, which is commonly exploited by scammers. The scammer attempts to create a striking headline that attracts the majority of users to click an attached link. Users who follow the link can be redirected to a fraudulent resource, where their personal data are easily extracted. To solve this problem, a novel browser extension named ClickBaitSecurity is proposed, which helps to evaluate the security of a link. The novel extension is based on the legitimate and illegitimate list search (LILS) algorithm and the domain rating check (DRC) algorithm. Both of these algorithms incorporate binary search features to detect malicious content more quickly and more efficiently. Furthermore, ClickBaitSecurity leverages the features of a deep recurrent neural network (RNN). The proposed ClickBaitSecurity solution has greater accuracy in detecting malicious and safe links compared to existing solutions.

2022 ◽  
Vol 2022 ◽  
pp. 1-11
Xiaorong Li ◽  
Lijun Zhang ◽  
Dongchen Li ◽  
Dan Guo

In this paper, RNN (Recurrent Neural Network) algorithm is used to conduct an in-depth analysis of HR strategic decision-making and an HR strategic decision model is constructed for simulation. The four evaluation dimensions of index screening are extracted, the generalized gray correlation analysis is applied to screen the initial selection indexes of HR strategic decision-making, and then the HR strategic decision-making index system of power supply enterprises is constructed. On this basis, the applicability of the BP neural network in HR strategic decision-making is analysed and demonstrated, a BP neural network-based HR strategic decision-making model for power supply enterprises is constructed, and the rationality of the model is illustrated through the model training and testing. Finally, an empirical study is conducted with S power supply company as an example to illustrate the operability of the BP network model for HR strategic decision-making. The results of this paper provide a scientific basis for human resource decision-making in power supply enterprises and provide theoretical support for promoting the healthy development of power supply enterprises. Improving the level of human resource management can make fuller use of human resources and realize the strategic goals of the enterprise. A strategy for optimizing the training of hybrid convolutional neural networks is proposed, using an exponential linear unit activation function to solve the problem of neuron dead zones, a hybrid pooling strategy is used to improve the problem of feature information loss in maximum pooling processing, and an improved cross-entropy loss function is used to solve the problem of insufficient learning of difficult classification samples. The optimization of the model training process is finally completed, and the recommendation quality is improved. The Bayesian probability distribution table is learned to be filled in the expert data set; finally, the accuracy and effectiveness of the model are tested on the experimental platform. The results of the experiments show that the established Bayesian model beats the platform’s built-in intelligent method with a 78.2% win rate. That is, the model can make intelligent recommendations for strategies in staffed and unmanned platform decision-making and command-and-control combat units to execute tactical actions to achieve the best operational effectiveness.

PeerJ ◽  
2022 ◽  
Vol 10 ◽  
pp. e12752
Ryan S. Alcantara ◽  
W. Brent Edwards ◽  
Guillaume Y. Millet ◽  
Alena M. Grabowski

Background Ground reaction forces (GRFs) are important for understanding human movement, but their measurement is generally limited to a laboratory environment. Previous studies have used neural networks to predict GRF waveforms during running from wearable device data, but these predictions are limited to the stance phase of level-ground running. A method of predicting the normal (perpendicular to running surface) GRF waveform using wearable devices across a range of running speeds and slopes could allow researchers and clinicians to predict kinetic and kinematic variables outside the laboratory environment. Purpose We sought to develop a recurrent neural network capable of predicting continuous normal (perpendicular to surface) GRFs across a range of running speeds and slopes from accelerometer data. Methods Nineteen subjects ran on a force-measuring treadmill at five slopes (0°, ±5°, ±10°) and three speeds (2.5, 3.33, 4.17 m/s) per slope with sacral- and shoe-mounted accelerometers. We then trained a recurrent neural network to predict normal GRF waveforms frame-by-frame. The predicted versus measured GRF waveforms had an average ± SD RMSE of 0.16 ± 0.04 BW and relative RMSE of 6.4 ± 1.5% across all conditions and subjects. Results The recurrent neural network predicted continuous normal GRF waveforms across a range of running speeds and slopes with greater accuracy than neural networks implemented in previous studies. This approach may facilitate predictions of biomechanical variables outside the laboratory in near real-time and improves the accuracy of quantifying and monitoring external forces experienced by the body when running.

Sign in / Sign up

Export Citation Format

Share Document