scispace - formally typeset
Search or ask a question
Topic

Hybrid neural network

About: Hybrid neural network is a research topic. Over the lifetime, 1305 publications have been published within this topic receiving 18223 citations.


Papers
More filters
Journal ArticleDOI
TL;DR: A hybrid neural network model is proposed to extract entities and their relationships without any handcrafted features to achieve the state-of-the-art results on entity and relation extraction task.

200 citations

Journal ArticleDOI
01 Feb 2018
TL;DR: A new forecast approach based on combination of a neural network with a metaheuristic algorithm as the hybrid forecasting engine and a 2‐stage feature selection filter based on the information‐theoretic criteria of mutual information and interaction gain, which filters out the ineffective input features is proposed.
Abstract: Prediction of solar power involves the knowledge of the sun , atmosphere and other parameters, and the scattering processes and the specifications of a solar energy plant that employs the sun's energy to generate solar power . This prediction result is essential for an efficient use of the solar power plant, the management of the electricity grid, and solar energy trading. However, because of nonlinear and nonstationary behavior of solar power time series, an efficient forecasting model is needed to predict it. Accordingly, in this paper, we propose a new forecast approach based on combination of a neural network with a metaheuristic algorithm as the hybrid forecasting engine. The metaheuristic algorithm optimizes the free parameters of the neural network. This approach also includes a 2‐stage feature selection filter based on the information‐theoretic criteria of mutual information and interaction gain, which filters out the ineffective input features. To demonstrate the effectiveness of the proposed forecast approach, it is implemented on a real‐world engineering test case. Obtained results illustrate the superiority of the proposed approach in comparison with other prediction methods.

189 citations

Proceedings ArticleDOI
Yilong Yang1, Qingfeng Wu1, Ming Qiu1, Yingdong Wang1, Xiaowei Chen1 
08 Jul 2018
TL;DR: Results indicate that the proposed pre-processing method can increase emotion recognition accuracy by 32% approximately and the model achieves a high performance with a mean accuracy of 90.80% and 91.03% on valence and arousal classification task respectively.
Abstract: As a challenging pattern recognition task, automatic real-time emotion recognition based on multi-channel EEG signals is becoming an important computer-aided method for emotion disorder diagnose in neurology and psychiatry. Traditional machine learning approaches require to design and extract various features from single or multiple channels based on comprehensive domain knowledge. Consequently, these approaches may be an obstacle for non-domain experts. On the contrast, deep learning approaches have been used successfully in many recent literatures to learn features and classify different types of data. In this paper, baseline signals are considered and a simple but effective pre-processing method has been proposed to improve the recognition accuracy. Meanwhile, a hybrid neural network which combines `Convolutional Neural Network (CNN)’ and `Recurrent Neural Network (RNN)’ has been applied to classify human emotion states by effectively learning compositional spatial-temporal representation of raw EEG streams. The CNN module is used to mine the inter-channel correlation among physically adjacent EEG signals by converting the chain-like EEG sequence into 2D-like frame sequence. The LSTM module is adopted to mine contextual information. Experiments are carried out in a segment-level emotion identification task, on the DEAP benchmarking dataset. Our experimental results indicate that the proposed pre-processing method can increase emotion recognition accuracy by 32% approximately and the model achieves a high performance with a mean accuracy of 90.80% and 91.03% on valence and arousal classification task respectively.

187 citations

Journal ArticleDOI
TL;DR: Thinker is an energy efficient reconfigurable hybrid-NN processor fabricated in 65-nm technology designed to exploit data reuse and guarantee parallel data access, which improves computing throughput and energy efficiency.
Abstract: Hybrid neural networks (hybrid-NNs) have been widely used and brought new challenges to NN processors. Thinker is an energy efficient reconfigurable hybrid-NN processor fabricated in 65-nm technology. To achieve high energy efficiency, three optimization techniques are proposed. First, each processing element (PE) supports bit-width adaptive computing to meet various bit-widths of neural layers, which raises computing throughput by 91% and improves energy efficiency by $1.93 \times $ on average. Second, PE array supports on-demand array partitioning and reconfiguration for processing different NNs in parallel, which results in 13.7% improvement of PE utilization and improves energy efficiency by $1.11 \times $ . Third, a fused data pattern-based multi-bank memory system is designed to exploit data reuse and guarantee parallel data access, which improves computing throughput and energy efficiency by $1.11 \times $ and $1.17 \times $ , respectively. Measurement results show that this processor achieves 5.09-TOPS/W energy efficiency at most.

185 citations

Journal ArticleDOI
TL;DR: A hybrid neural network that combines the advantages of a convolutional neural network with those of long short-term memory is designed for model training and prediction and demonstrates wide generality and reduced errors when compared with the other state-of-the-art methods.

181 citations


Network Information
Related Topics (5)
Artificial neural network
207K papers, 4.5M citations
89% related
Feature extraction
111.8K papers, 2.1M citations
88% related
Fuzzy logic
151.2K papers, 2.3M citations
85% related
Convolutional neural network
74.7K papers, 2M citations
84% related
Deep learning
79.8K papers, 2.1M citations
83% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
20233
20228
2021128
2020119
2019104
201863