2170 results about "Recurrent neural network" patented technology

The invention provides a machine reading understanding method based on a multi-head attention mechanism and dynamic iteration, and belongs to the field of natural language processing. The constructionmethod of the machine reading understanding model comprises the following steps: constructing an article and problem coding layer; constructing a recurrent neural network based on bidirectional attention flow; a self-attention layer is constructed and answer output is predicted based on a dynamic iterative decoder. According to the method, answer prediction can be carried out on questions in a machine reading understanding task text; according to the invention, a new end-to-end neural network model is established, and a new idea is provided for exploring a machine reading understanding task.

The invention relates to an electric power telecommunication network device fault prediction method based on improved LSTM. The invention provides a data preprocessing and time sequence input construction method for the first time. Compared with the simple recurrent neural network, the LSTM is more likely to learn long-term dependence and can solve the prediction problems related to sequences. Because of the strong association between device alerts, the independence of variables can be ensured through PCA. The target replication strategy is also used for improving the LSTM, which can bring local error information in every step. Compared with a simple target output only in the last step, the strategy can improve the accuracy of the model and reduce the risk of over fitting. In combination with dropout, the invention proposes a prediction model of LSTM, which can achieve better prediction precision by deep learning. At the same time, the LSTM is usd for modeling electric power telecommunication network alarm data for the first time and identifying the timing sequence mode therein.

The invention discloses an aspect-level emotion classification model and method based on dual-memory attention, belonging to the technical field of text emotion classification. The model of the invention mainly comprises three modules: an encoder composed of a standard GRU loop neural network, a GRU loop neural network decoder introducing a feedforward neural network attention layer and a Softmaxclassifier. The model treats input statements as a sequence, based on the attention paid to the position of the aspect-level words in the sentence, Two memory modules are constructed from the originaltext sequence and the hidden layer state of the encoder respectively. The randomly initialized attention distribution is fine-tuned through the attention layer of the feedforward neural network to capture the important emotional features in the sentences, and the encoder-decoder classification model is established based on the learning ability of the GRU loop neural network to the sequence to achieve aspect-level affective classification capabilities. The invention can remarkably improve the robustness of the text emotion classification and improve the classification accuracy.

The present invention relates to a control algorithm constructing device that constructs a control algorithm controlling the motion of a robot, and a controller that controls the motion of the robot in accordance with the constructed control algorithm, with the purpose of reducing the cost and time taken to create the control algorithm as compared with the conventional method such as an MZP method to solve a mechanical equation, in which the control algorithm is constructed by a recurrent neural network (RNN) including a neuron generating an output with an analogue lag with respect to an input, the coefficients of the RNN are determined in succession from the term of lower degree to the term of higher degree.

The invention discloses a question and answering (QA) system realization method based on deep learning and a topic model. The method comprises the steps of: S1, inputting a question sentence to the Twitter LDA topic model to obtain a topic type of the question sentence, extracting a corresponding topic word, and indicating the input question sentence and the topic word as word vectors; S2, inputting word vectors of the input question sentence to a recurrent neural network (RNN) for encoding to obtain an encoded hidden-layer state vector of the question sentence; S3, using a joint attention mechanism and combining local and global hybrid semantic vectors of the question sentence by a decoding recurrent neural network for decoding to generate words; S4, using a large-scale conversation corpus to train a deep-learning topic question and answering model based on an encoding-decoding framework; and S5, using the trained question and answering model to predict an answer to the input questionsentence, and generating answers related to a question sentence topic. The method makes up for the lack of exogenous knowledge of question and answering models, and increases richness and diversity of answers.

The invention provides a natural scene text detection method and system. The system comprises two neural network models: a text detection network based on multi-level semantic feature fusion and a detection screening network based on an attention mechanism, wherein the text detection network is an FCN-based image feature extraction fusion network, and is used for extracting multi-semantic level information of input data, carrying out full fusion of multi-scale features, and finally predicting the position and confidence degree of text information in a natural scene by carrying out convolutionoperation on the fused multi-scale information. According to the detection screening network, the trained convolutional recurrent neural network is used for discriminating and scoring an initial detection result output by the convolutional neural network of the first part, so that a background which is easy to confuse with foreground characters is filtered out, and the accuracy of natural scene text recognition is further improved.

The invention discloses an online traditional Chinese medicine text named entity identifying method based on deep learning. The method includes the steps that online traditional Chinese medicine text data are obtained through a web crawler, and named entities of the obtained online traditional Chinese medicine text data are labeled with existing terminological dictionaries and human assistance; a word2vec tool is used for carrying out learning on large-scale label-free linguistic data, and word vectors with fixed length are obtained and used for forming a corresponding glossary; word segmentation is carried out on the online traditional Chinese medicine text data, words are converted into the word vectors with the fixed length by searching for the glossary, the word vectors serve as input of a convolutional neural network, and a blank character is used for filling when sentence length is insufficient; output of the convolutional neural network serves as input of a bidirectional long-short-time memory recurrent neural network, and an identification result of the online traditional Chinese medicine text data words to be identified is output. Compared with a traditional method for named entity identifying, the method reduces complexity and workload of feature extraction, simplifies the processing process and remarkably improves identification efficiency.

The invention provides a dialogue data interaction processing method based on a recurrent neural network. The method comprises the following steps: receiving a dialogue input statement of a user; carrying out knowledge base matching calculation, and judging whether a knowledge base has a problem statement, the matching degree between which and the dialogue input statement reaches a preset value; if not, requesting a dialogue generation model to give an answer to the dialogue input statement, wherein a coding layer of the dialogue generation model is constructed into the recurrent neural network, analyzing the dialogue input statement in the recurrent neural network to obtain a middle vector expressing problem meanings, and a decoding layer of the dialogue generation model is also constructed into the recurrent neural network, analyzing the middle vector in the recurrent neural network to obtain an answer vector group expressing answer meanings; and taking the answer vector group as an answer output statement and outputting the answer output statement. The method can enable human-machine interaction to be smoother; and answers given by the dialogue generation model can further enable the knowledge base to be expanded and updated.

The invention discloses a short text emotion factor extraction method and device based on deep learning and relates to the technical field of machine translation, aiming at solving the technical problems that the understanding on natural languages of an existing computer system is still at a relatively low phase and an identification result of an emotion evaluation object is not good. According to the technical scheme, the short text emotion factor extraction method comprises: modeling sentences through adopting a recurrent neural network of bidirectional long-time and short-time memory; constructing a classifier for each type and classifying; with regard to input sentences, taking each word in the sentences as a word vector, taking the word vector as an input sequence and inputting the word vector into the recurrent neural network; calculating each hide status in the recurrent neural network in sequence and calculating character representation of the sentences; and after obtaining the character representation of the sentences, classifying the sentences by adopting a logic classifier and identifying types of emotion factor commented in the sentences.

The invention discloses a text classification method based on a bidirectional cyclic attention neural network, and belongs to the technical field of learning and natural language processing. The method comprises the following steps: step 1, preprocessing data; Step 2, according to the preprocessed data, generating and training a word vector of each word through a Word2vec method; Step 3, performing text semantic feature extraction on the word vector according to the word vector, fusing an attention mechanism and a bidirectional recurrent neural network, calculating the overall weight of each word, and converting the weight into an output value Y (4) of a model; And step 4, taking the feature vector Y (4) as the input of a softmax classifier according to the feature vector Y (4), and carrying out classification identification. According to the method, the attention mechanism is fused in the text feature learning model, the effect of keywords can be effectively highlighted, the performance of the model is greatly improved, and the text classification accuracy is further improved.

The invention relates to a psychological assessment method based on a virtual reality technology. Virtual scenarizing processing is carried out on a psychological scale; answers, behaviors, and physiological data of a subject are collected in real time; on the basis of the answer options of the subject, intelligent jumping of problems of the psychological scale is completed; the scale content is analyzed comprehensively and intelligently, three kinds of data are trained by a convolution neural network and a recurrent neural network and then feature fusion is carried out, the processed information is inputted into a softmax layer to obtain a psychological assessment model; with the psychological assessment model, the psychological assessment result of the subject is compared with a doctor tag, loss function calculation and gradient reversed conduction are carried out to correct the answer options of the subject intelligent; and then the psychological, behavior and the corrected answer data are calculated by the psychological assessment model to obtain a final assessment result. According to the invention, on the basis of combination of VR, intelligent sensing, big data analysis, artificial intelligence and other technologies with the traditional psychological assessment method, the accuracy of psychological assessment is improved and the medical resources are saved effectively.

The invention provides a face image completion method based on a self-attention deep generative adversarial network. The method comprises the following steps: constructing a face image completion model comprising an attention recurrent neural network module, a generator network and a discriminator network; carrying out data collection: collecting a large number of face images to form an image set, and dividing the image set into a training set and a test set; carrying out image preprocessing to enable the size of the image to be suitable for processing in a deep learning network; constructing a damaged face image serving as model input; carrying out model training, using a GAN framework and various regularization means to directly train the generator network and the discriminator network at the same time in an end-to-end mode, and when the generator network and the discriminator network reach theoretical Nash equilibrium, completing model training; training the model. Compared with the prior art, the face image complementing method based on the self-attention deep generative adversarial network has the advantage that the complemented face image is better in quality.

The invention discloses an electromyographic signal gesture recognition method based on deep learning and attention mechanisms. The method comprises the following steps: performing noise reduction filtering on electromyographic signals; extracting one classic characteristic set from each wind datum by using a sliding window, and establishing a new electromyographic image based on characteristics;designing a deep learning frame based on a convolutional neural network, a circulation neural network and an attention mechanisms, and optimizing network structure parameters of the deep learning frame; performing training with the designed deep learning frame and the training data so as to obtain a classifier model; inputting testing data into the trained deep learning network model, and according to likelihood of a last layer of output, maximally likelihooding corresponding types, that is, recognition types. By adopting the method, electromyographic gesture signals can be recognized on the basis of new characteristic images and deep learning frames based on attention mechanisms. By adopting the electromyographic signal gesture recognition method based on deep learning and attention mechanisms, multiple different gestures of a same subject can be accurately recognized.

An embodiment of the invention provides a searching-ranking model training method and device and a search processing method. The searching-ranking model training method includes: acquiring multiple groups of labeled sample data, wherein each group of the sample data includes search terms and multiple search result entries correspondingly labeled as positive examples or negative examples; generating Gated RNN (recurrent neural network) based input layer, word vector layer, hide layer and output layer of a searching-ranking model, and training the searching-ranking model to learn parameters of the same. By the searching-ranking model training method and device and the search processing method, accuracy of ranking scores of the search result entries acquired by calculating can be improved, and more accurate searching-ranking results are provided for users.

Disclosed is a deep long-term and short-term memory recurrent neural network acoustic model establishing method based on selective attention principles. According to the deep long-term and short-term memory recurrent neural network acoustic model establishing method based on the selective attention principles, attention gate units are added inside a deep long-term and short-term memory recurrent neural network acoustic model to represent instantaneous function change of auditory cortex neurons; the gate units are different in other gate units in that the other gate units are in one-to-one correspondence with time series, while the attention gate units represent short-term plasticity effects and accordingly have intervals in the time series; through the neural network acoustic model obtained by training mass voice data containing Cross-talk noise, robustness feature extraction of the Cross-talk noise and establishment of robust acoustic models can be achieved; the aim of improving the robustness of the acoustic models can be achieve by inhibiting influence of non-target flow on feature extraction. The deep long-term and short-term memory recurrent neural network acoustic model establishing method based on the selective attention principles can be widely applied to multiple voice recognition-related machine learning fields of speaker recognition, keyword recognition, man-machine interaction and the like.

The method is suitable for the technical field of traffic prediction, and provides a traffic flow prediction method and system, and a terminal device. The method comprises the following steps of: acquiring traffic data in a specified time period of a target area, performing graph convolution network training on the traffic data, capturing a topological structure of an urban road network to obtainspatial features, inputting the time sequence of the spatial features into a gating cycle unit model, and obtaining dynamic changes through information transmission between units to capture time features; and establishing a prediction model according to the spatial features and the time features so as to predict the traffic flow of the target area according to the prediction model. According to the invention, a graph convolutional neural network based on an attention mechanism is applied to the prediction problem in the traffic field, high-dimensional features in the road traffic network can be better mined in a complex road structure, the accuracy of traffic flow prediction is improved in cooperation with the improved recurrent neural network, and the improvement effect is obvious especially in medium-term and long-term prediction.

The invention discloses a multiple features fused bidirectional recurrent neural network fine granularity opinion mining method. The method comprises the following steps of: capturing comment data of a specific website through internet and carrying out labelling and preprocessing on the comment data to obtain a training sample set; carrying out training by using a Word2Vec or Glove model algorithm to obtain word vectors of the comment data; carrying out vectorization after carrying out part of speech labeling, dependence relationship labeling and the like; and inputting the vectors into a bidirectional concurrent neural network to construct a bidirectional recurrent neural network fine granularity opinion mining model. According to the method, attribute words in fine granularity opinion mining is extracted and emotional polarity judgement is carried out through the training of a model, so that plenty of model training time is further saved and the training efficiency is improved; no professionals are required to carry out manual extraction on the attribute words, so that a lot of manpower cost is saved; and moreover, the model can be trained by using a plurality of data sources, so that cross-field fine granularity opinion analysis can be completed, thereby solving the problem of long-distance emotional element dependency.

The invention discloses a text label marking device, which is used for marking a text label. The device comprises an input module suitable to receive text inputs and convert and output texts to a vector matrix, a convolutional neural network module connected with the input module and suitable to output local semantic features of the texts according to the vector matrix, a recurrent neural networkmodule connected with the input module and suitable to output long-distance semantic features of the texts according to the vector matrix, an attention model module connected with the convolutional neural network module and the recurrent neural network module and suitable to output weights of individual characters in the texts according to the local semantic features and the long-distance semanticfeatures, and an output module connected with the attention model module and suitable to receive the weights of the individual characters in the texts and output text labels and probabilities of thelabels. The invention furthermore discloses a training method for the text label marking device, a corresponding text label marking method and a computing device.

The invention relates to a reservoir prediction method based on a bidirectional recurrent neural network. The reservoir prediction method includes: generating well seismic data, generating well markeddata, generating a marked dataset, building and training the bidirectional recurrent neural network, storing a trained model, extracting seismic data corresponding to all geographic positions in a region, and performing prediction on the seismic data to acquire probability distribution of reservoirs below the whole region. A one-to-one corresponding relation between each sampling point and reservoirs and non-reservoirs is established, wherein input of each time step is n-dimension seismic data of each sampling point while output of the same is corresponding reservoir or non-reservoir marks. An optimal model is acquired by training marked samples and adjusting hyper-parameters, in this way, unmarked data can be predicted to finally generate a predicted seismic body of same dimension, and value of each sampling point is prediction probability between 0 and 1, so that the method has good effect on reservoir prediction.

Systems and methods are provided to identify abnormal transaction activity by a participant that is inconsistent with current conditions. Historical participant and external data is identified. A recurrent neural network identifies patterns in the historical participant and external data. A new transaction by the participant is received. The new transaction is compared using the patterns to the historical participant and external data. An abnormality score is generated. An alert is generated if the abnormality score exceeds a threshold.