Article
Biochemical Research Methods
Ngoc-Quang Nguyen, Gwanghoon Jang, Hajung Kim, Jaewoo Kang
Summary: Compound-protein interaction (CPI) is crucial in drug discovery, and there have been AI-based approaches proposed to study it. Two types of models, graph convolutional neural networks and neural networks applied to molecular descriptors or fingerprints, have shown promising results. However, it is still unclear which method is superior. This study presents the Perceiver CPI network, which utilizes a cross-attention mechanism and rich information from extended-connectivity fingerprints to enhance the learning ability and performance. The proposed method outperforms previous approaches in all experiments on three main datasets.
Article
Automation & Control Systems
Ning Jin, Yongkang Zeng, Ke Yan, Zhiwei Ji
Summary: Artificial intelligence-based air quality index (AQI) forecasting is a hot research topic, and the proposed multiple nested long short term memory network (MTMC-NLSTM) model performs superior in accurate AQI forecasting.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Energy & Fuels
Hui Song, Nameer Al Khafaf, Ammar Kamoona, Samaneh Sadat Sajjadi, Ali Moradi Amani, Mahdi Jalili, Xinghuo Yu, Peter McTaggart
Summary: With the increasing importance of renewable energy, predicting photovoltaic (PV) power generation becomes crucial for power management and optimization. This paper proposes a multitasking prediction approach using recurrent neural networks (RNNs) to improve the accuracy of PV power generation prediction across different customer categories. The proposed multitasking RNN (MT-RNN) framework transfers knowledge among tasks, achieving superior performance compared to individual deep neural network (DNN) models.
Article
Computer Science, Artificial Intelligence
Marcos Eduardo Valle, Rodolfo Anibal Lobo
Summary: The researchers extended bipolar RCNNs to deal with hypercomplex-valued data and investigated the stability of these new networks. Examples were provided to illustrate the theoretical results and computational experiments confirmed the potential application of hypercomplex-valued RCNNs as associative memories for gray-scale images.
Article
Multidisciplinary Sciences
Bruno B. Averbeck
Summary: Adolescent development is marked by improved cognitive processes, but a decline in the ability to learn new skills. Pruning of synapses occurs during this period. Our study shows that pruned neural networks perform better on certain tasks but learn new problems more slowly, indicating that overproduction and subsequent pruning of synapses is a computationally advantageous approach to developing a competent brain.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2022)
Article
Thermodynamics
Jikai Duan, Hongchao Zuo, Yulong Bai, Jizheng Duan, Mingheng Chang, Bolong Chen
Summary: This study presents a novel hybrid forecasting system that significantly enhances the accuracy of wind speed prediction, achieving superior prediction results compared to other models through a process of decomposition, prediction, and error correction.
Article
Chemistry, Medicinal
Xin Lai, Peisong Yang, Kunfeng Wang, Qingyuan Yang, Duli Yu
Summary: MGRNN is a neural network model for drug molecular structure generation with features including efficient training computation, data robustness, and iterative sampling. Experimental results demonstrate MGRNN's ability to generate a high percentage of chemically valid molecules even without chemical knowledge.
MOLECULAR INFORMATICS
(2021)
Article
Chemistry, Multidisciplinary
Dejun Jiang, Zhaofeng Ye, Chang-Yu Hsieh, Ziyi Yang, Xujun Zhang, Yu Kang, Hongyan Du, Zhenxing Wu, Jike Wang, Yundian Zeng, Haotian Zhang, Xiaorui Wang, Mingyang Wang, Xiaojun Yao, Shengyu Zhang, Jian Wu, Tingjun Hou
Summary: This study compiled the largest metalloprotein-ligand complex dataset and evaluated the docking powers of three competitive docking tools for metalloproteins. A structure-based deep graph model called MetalProGNet was developed to predict metalloprotein-ligand interactions. MetalProGNet outperformed various baselines in internal and external evaluations. The study also employed an atom-atom interaction masking technique to interpret MetalProGNet.
Article
Computer Science, Artificial Intelligence
Yufei Cui, Yu Mao, Ziquan Liu, Qiao Li, Antoni B. Chan, Xue Liu, Tei-Wei Kuo, Chun Jason Xue
Summary: Nested dropout is a variant of dropout operation that orders network parameters or features based on pre-defined importance. It has been explored in constructing nested nets and learning ordered representation.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Biology
Yue Kris Wu, Friedemann Zenke
Summary: Neural circuits can achieve rapid information processing through nonlinear transient amplification, which involves two phases - selective amplification of inputs exceeding a critical threshold by positive feedback excitation, and stabilization of runaway dynamics into an inhibitory state by short-term plasticity. NTA offers a parsimonious explanation for how excitatory-inhibitory co-tuning and short-term plasticity collaborate in recurrent networks to achieve transient amplification.
Article
Chemistry, Medicinal
Jian Wang, Nikolay Dokholyan
Summary: This study finds that deep learning-based methods for predicting binding affinities lack generalizability, and a newly developed predictor, Yuel, shows better ability in predicting interactions between unknown compounds and proteins.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2022)
Article
Computer Science, Artificial Intelligence
Venkateswarlu Gundu, Sishaj P. Simon
Summary: This paper proposes an accurate forecasting model for solar energy and temperature based on LSTM, obtaining a suitable network structure through statistical analysis. The model shows higher prediction accuracy for solar and temperature data compared to traditional network models available in the literature.
NEURAL PROCESSING LETTERS
(2021)
Article
Computer Science, Information Systems
Madiha Bukhsh, Muhammad Saqib Ali, Muhammad Usman Ashraf, Khalid Alsubhi, Weiqiu Chen
Summary: This paper presents a flexible method for interpreting the relational structure between polynomial roots and coefficients using LSTM-RNN. An adaptive learning optimization algorithm is used to prevent weight fluctuation. Experimental results show the superiority of the proposed LSTM-RNN model in approximating polynomial roots.
Article
Computer Science, Artificial Intelligence
R. Dharaniya, J. Indumathi, G. V. Uma
Summary: The objective of this study is to perform text generation specifically for movie scripts, identifying context and building scripts through sentiment classification and text vectorization. Bidirectional long short-term memory and multi-head attention mechanism are used to understand future context.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Venkateswarlu Gundu, Sishaj P. Simon, Krishna Kumba
Summary: In this study, multiple deep neural network models and different weather parameters are used to forecast solar power generation. Through comparing the performance of these models and testing with actual data, it is found that the forecast model based on JAYA-based LSMN demonstrates superior predicting performance compared to conventional techniques.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)