Article
Engineering, Chemical
Jun Zhang, Qin Wang, Yang Su, Saimeng Jin, Jingzheng Ren, Mario Eden, Weifeng Shen
Summary: This study developed an accurate and interpretable deep neural network (AI-DNN) model for predicting lipophilicity. A hybrid method of molecular representation, combining directed message passing neural networks and fixed molecule-level features, was employed to capture the local and global features of molecules. The proposed model demonstrated promising predictive accuracy and discriminative power in structural and stereoisomers. The use of Monte Carlo Tree Search allowed for interpretation of the model, which is important in fields with a high demand for interpretable deep networks, such as green solvent design and drug discovery.
Article
Chemistry, Multidisciplinary
Qingchang Liu, Yuan Gao, Baoxing Xu
Summary: The study successfully predicted the thermal conductivity of graphene structures under external mechanical loading using a deep neural network-based machine learning method, providing a foundation for establishing mechanically adaptive structures and responsive thermal property paradigms.
Review
Computer Science, Artificial Intelligence
Weiwei Jiang
Summary: This article provides a latest review of deep learning models for stock market prediction, categorizing data sources, neural network structures, and evaluation metrics to help researchers stay updated and easily reproduce previous studies. It also highlights some future research directions in this topic.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Review
Biochemical Research Methods
Karin Elimelech-Zohar, Yaron Orenstein
Summary: Nucleic-acid G-quadruplexes (G4s) are crucial in cellular processes, and experimental assays have been developed to measure them in high throughput. This has enabled the development of machine-learning-based methods, particularly deep neural networks, to predict G4s in any nucleic-acid sequence and species.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Health Care Sciences & Services
Chang Seok Bang, Hyun Lim, Hae Min Jeong, Sung Hyeon Hwang
Summary: This study demonstrates the usefulness of AutoDL in establishing customized deep learning models on-site. An inexperienced endoscopist with a certain level of expertise can benefit from the support of AutoDL.
JOURNAL OF MEDICAL INTERNET RESEARCH
(2021)
Article
Engineering, Multidisciplinary
Zi-hang Yang, Ji-li Rong, Zi-tong Zhao
Summary: The accurate and efficient prediction of explosive detonation properties is of great significance for weapon design, and recent advances in deep learning methods have allowed researchers to apply them to this field. However, there are still challenges in deep learning-based studies of detonation properties. This work establishes three improved deep learning models and tests their accuracy, demonstrating their utility in engineering applications.
DEFENCE TECHNOLOGY
(2023)
Article
Thermodynamics
Lei Sun, Tianyuan Liu, Yonghui Xie, Di Zhang, Xinlei Xia
Summary: Deep learning models, particularly the RNN model, show superior performance in balancing accuracy and efficiency for turbine power prediction compared to shallow models and typical machine learning models. The study also highlights the influence of training size and input time-steps on the performance of the RNN model, demonstrating its potential in improving the accuracy of turbine power prediction.
Article
Environmental Sciences
Rajnish Kumar, Farhat Ullah Khan, Anju Sharma, Mohammed Haris Siddiqui, Izzatdin B. A. Aziz, Mohammad Amjad Kamal, Ghulam Md Ashraf, Badrah S. Alghamdi, Md Sahab Uddin
Summary: Identifying and screening mutagenic chemicals are crucial for ensuring the safety of chemical compounds, and prediction models developed using machine learning techniques such as DNN have shown higher accuracy and performance.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2021)
Article
Engineering, Chemical
Huaqiang Wen, Yang Su, Zihao Wang, Saimeng Jin, Jingzheng Ren, Weifeng Shen, Mario Eden
Summary: The research proposes a systematic approach to solving key problems in DNN-based QSPR modeling, including applicability domain and prediction uncertainty, using multiple machine learning technologies. The method extracts features through principal component analysis and kernel PCA, defines a detailed applicability domain using the K-means algorithm, and further analyzes prediction uncertainty.
Article
Computer Science, Artificial Intelligence
Yongxiang Huang, Albert C. S. Chung
Summary: In this paper, a graph-convolutional framework for population-based disease prediction on multi-modal medical data is proposed. The framework can automatically learn to build a population graph with variational edges and substantially improve the predictive accuracy by incorporating multi-modal data.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Computer Science, Information Systems
Subba Rao Polamuri, Kudipudi Srinivas, A. Krishna Mohan
Summary: Deep learning has achieved significant success in the financial domain, particularly in stock market prediction. This paper addresses the limited use of generative adversarial networks (GANs) in stock market prediction by proposing a GAN-based hybrid prediction algorithm. The algorithm overcomes the difficulty in setting hyperparameters using reinforcement learning and Bayesian optimization. Empirical results demonstrate the promising performance of the GAN-based deep learning framework (Stock-GAN) compared to the state-of-the-art model (MM-HPA) in stock price prediction.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Engineering, Environmental
Zeren Jiao, Chenxi Ji, Yue Sun, Yizhi Hong, Qingsheng Wang
Summary: In this study, a toxic dispersion database is constructed using PHAST simulations, and a quantitative consequence prediction model is developed using deep neural network algorithm, which shows the highest accuracy for predicting toxic dispersion downwind distances. These models can provide instant toxic dispersion range estimations for various toxic chemicals at lower computational costs.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2021)
Article
Chemistry, Analytical
Xinyao Feng, Ehsan Ahvar, Gyu Myoung Lee
Summary: This paper defines a smart home use case for adjusting home temperature and hot water automatically. The goal is to reduce energy consumption in smart homes by allowing residents to set their preferred temperature and automatically adjusting it based on their arrival time. The paper focuses on predicting residents' arrival time using machine learning-based street traffic prediction system and finding the most appropriate algorithm for integration. The evaluation is done using an Uber's dataset for San Francisco and the system can also be used as a route recommender.
Article
Computer Science, Artificial Intelligence
Hojjat Rakhshani, Lhassane Idoumghar, Soheila Ghambari, Julien Lepagnot, Mathieu Brevilliers
Summary: The study focuses on the performance of deep learning models in dealing with global optimization problems, and analyzes it through empirical experiments on CEC 2017 benchmark suite and protein structure prediction (PSP) problems. The results reveal that the generated learning models can achieve competitive results given enough computational budget.
APPLIED SOFT COMPUTING
(2021)
Article
Spectroscopy
Alexander A. Ksenofontov, Michail M. Lukanov, Pavel S. Bocharov, Michail B. Berezin, Igor Tetko
Summary: A new model based on a large dataset and deep neural network architecture was developed to accurately predict the absorption maximum wavelength of BODIPYs. The consensus model achieved higher accuracy and can accelerate the design of new dyes with desired properties.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2022)
Article
Polymer Science
Luis A. Miccio, Jon Otegui, Marcela E. Penoff, Pablo E. Montemartini, Gustavo A. Schwartz
JOURNAL OF APPLIED POLYMER SCIENCE
(2015)
Article
Chemistry, Physical
Manuel Monasterio, Juan J. Gaitero, Edurne Erkizia, Ana M. Guerrero Bustos, Luis A. Miccio, Jorge S. Dolado, Silvina Cerveny
JOURNAL OF COLLOID AND INTERFACE SCIENCE
(2015)
Article
Polymer Science
Jon Otegui, Luis A. Miccio, Arantxa Arbe, Gustavo A. Schwartz, Mathias Meyer, Stephan Westermann
RUBBER CHEMISTRY AND TECHNOLOGY
(2015)
Article
Chemistry, Multidisciplinary
Guillaume Vasseur, Mikel Abadia, Luis A. Miccio, Jens Brede, Aran Garcia-Lekue, Dimas G. de Oteyza, Celia Rogero, Jorge Lobo-Checa, J. Enrique Ortega
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
(2016)
Article
Chemistry, Multidisciplinary
Luis A. Miccio, Martin Setvin, Moritz Mueller, Mikel Abadia, Ignacio Piquero, Jorge Lobo-Checa, Frederik Schiller, Celia Rogero, Michael Schmid, Daniel Sanchez-Portal, Ulrike Diebold, J. Enrique Ortega
Article
Chemistry, Physical
Sara Blomberg, Johan Zetterberg, Jianfeng Zhou, Lindsay R. Merte, Johan Gustafson, Mikhail Shipilin, Adriana Trinchero, Luis A. Miccio, Ana Magana, Maxim Ilyn, Frederik Schiller, J. Enrique Ortega, Florian Bertram, Henrik Gronbeck, Edvin Lundgren
Article
Polymer Science
Guillermina Capiel, Luis A. Miccio, Pablo E. Montemartini, Gustavo A. Schwartz
POLYMER DEGRADATION AND STABILITY
(2017)
Article
Chemistry, Physical
Daniel E. Martinez-Tong, Luis A. Miccio, Angel Alegria
Article
Physics, Applied
Luis A. Miccio, Mohammed M. Kummali, Gustavo A. Schwartz, Angel Alegria, Juan Colmenero
JOURNAL OF APPLIED PHYSICS
(2014)
Article
Polymer Science
Luis A. Miccio, Gustavo A. Schwartz
Article
Polymer Science
Luis A. Miccio, Gustavo A. Schwartz
Summary: This work explores the use of artificial neural networks to predict polymer properties and encode chemical structures. By embedding monomer chemical structures in a high-dimensional abstract space and employing neural network training and clustering methods, accurate prediction of polymer properties and structure encoding were successfully achieved.
Article
Polymer Science
Luis A. Miccio, Claudia Borredon, Ulises Casado, Anh D. Phan, Gustavo A. Schwartz
Summary: The analysis of structural relaxation dynamics of polymers provides insights into their mechanical properties, which are important for determining a material's suitability for practical applications. However, obtaining the relaxation time through experimental processes after polymer synthesis is time-consuming. In this study, we propose a method that combines artificial neural networks and the elastically collective nonlinear Langevin equation (ECNLE) to estimate the temperature dependence of the main structural relaxation time of polymers based solely on the chemical structure of the monomer.
Article
Mathematics, Interdisciplinary Applications
Luis A. Miccio, Carlos Gamez-Perez, Juan Luis Suarez, Gustavo A. Schwartz
Summary: This study uses complex network analysis to examine the role of social and cultural networks in the emergence of prominent historical figures and ideas. By analyzing the interactions in Wikipedia, the researchers discovered new knowledge about the interdisciplinary transactions between individuals associated with the Italian Renaissance. The findings highlight the strong network-level interactions between certain individuals, the importance of specific ideas across different clusters, and the influence of knowledge dealers on historical depictions of the period. The study also introduces the concept of focus reading, which utilizes complex network analysis to bridge different forms of historical evidence.
ADVANCES IN COMPLEX SYSTEMS
(2022)
Proceedings Paper
Physics, Applied
Luis A. Miccio, Gustavo A. Schwartz
TIMES OF POLYMERS (TOP) AND COMPOSITES 2014
(2014)
Article
Polymer Science
Yuanzhang Jiang, Xuyi Wang, Yanting Han, Dakai Gong, Yingchun Gu, Lin Tan
Summary: In this study, FeCl3 and poly(acrylic acid) (PAA) were doped into waterborne polyurethanes (WPUs) to create multifunctional materials with self-healing and antimicrobial properties. WPU/Fe/PAA-5.5, containing 5.5 wt % PAA, exhibited excellent mechanical properties, strong self-healing capabilities, and an impressive antimicrobial rate of over 90%. These supramolecular films were also highly recyclable through hot-pressing.
Article
Polymer Science
Longyu Tian, Min Wang, Guangming Liao, Baoliang Liu, Sujuan Zhang, Yucheng Sun, Zhen Meng, Jintao Zhang, Zaijun Lu
Summary: Two kinds of benzoxazine monomers were synthesized and used to prepare polybenzoxazine anion exchange membranes (AEMs) with quaternary ammonium functionalization. The LQPBZ AEM exhibited better ion conductivity and alkaline stability.
Article
Polymer Science
Claire A. Lemarchand
Summary: This study investigates the shock behavior of three different polymers through molecular dynamics simulations. The results reveal specificities in the shock behavior of polymers, including deviations from the linear relation between shock velocity and particle velocity, as well as differences in shear stress relaxation behind the shock front. It is found that the deviation of the Hugoniot locus is related to the change in the relative contribution of bonding and non-bonding potential energies, while polymers with higher glass transition temperatures exhibit slower shear stress relaxation.
Article
Polymer Science
Zengquan Liu, Xiaochun Yin, He Zhang, Shuo Gao, Qinglin Kuang, Yanhong Feng
Summary: A powder solid-state extension (PSSE) technology was proposed to address the high melt viscosity issue of UHMWPE. By uniaxial extension and sintering in the solid state, a highly self-reinforced material was prepared. X-ray diffraction and scattering techniques were used to study the structural evolution during the PSSE process, and a solid-phase induced transformation model was established.
Article
Polymer Science
Zhike Li, Anyu Luo, Rui Zhou, Xin Li, Haiyan Li
Summary: A novel high temperature resistance IL@SiO2 nanocapsules were successfully prepared and their friction and wear properties in PA6 composites were studied. The addition of IL@SiO2 nanocapsules reduced the friction coefficient and wear rate without affecting the mechanical properties of PA6. The synergy between the IL core and SiO2 wall in the nanocapsules improved the self-lubricating performance of the PA6 composites.
Article
Polymer Science
Xin Liu, Xuhong Guo, Qi Liao
Summary: In this study, an algorithm for accurately estimating viscosity is developed using molecular dynamics simulations and the Green-Kubo formula. This algorithm can be applied to complex systems with long correlations, such as macromolecular and biological simulation systems.
Article
Polymer Science
Hao Pu, Yun-Lei Hou, Jing -Zhou Chen, Dong -Lin Zhao
Summary: The use of modified graphene improves the interfacial adherence of carbon fibers to epoxy resin, resulting in enhanced interfacial and bending properties of CF and matrix in CF/EP composites, as well as increased interlaminar shear strength and flexural strength.
Article
Polymer Science
Wei-Chung Ke, Jin-Wei Lin, Manohar Reddy Busireddy, Yueh-Hsing Lee, Jiun-Tai Chen, Chain-Shu Hsu
Summary: This study introduces a crosslinkable monomer, TAIC, to synthesize three crosslinked polyimide films, which exhibit improved thermal, mechanical, and dielectric properties. The introduction of TAIC enhances the tensile strength, reduces the dielectric constant and dielectric loss, and decreases the coefficient of thermal expansion of the polyimide films.
Article
Polymer Science
Alex Kwasi Kumi, Ruiling Fan, Ye Chen, Yumei Zhang
Summary: The difference in leaching of amylopectin from cellulose/amylopectin/1-butyl-3-methylimidazolium chloride ([Bmim][Cl]) blends during regeneration in water and aqueous ethanol has been studied. Molecular dynamics simulations showed that the dissolution and regeneration mechanisms of amylopectin in [Bmim][Cl] are similar to cellulose in ionic liquids. Water regeneration leads to weak electrostatic interactions, resulting in high leaching of amylopectin. In contrast, ethanol-water interactions enhance electrostatic interactions among amylopectin chains, limiting amylopectin leaching in aqueous ethanol.
Article
Polymer Science
Xi Zeng, Junwei Zhou, Junbiao Peng, Chunsheng Zhang, Danling Wang, Yihu Song, Qiang Zheng
Summary: This study utilizes an oligomer deep eutectic solvent (DES) based on polyethylene glycol to regulate the vulcanization kinetics of rubber. The results show that the DES can accelerate vulcanization at low temperatures without affecting the crosslinking density and Mullins effect.
Article
Polymer Science
Qianqian Yue, Yongfei Peng, Xingjian Liu, Aihua He, Huarong Nie
Summary: The addition of metal deactivators can effectively suppress the aging of TBIR, prolong its lifespan, and have no negative effects on its physical and processing properties. These research findings are of great importance for improving the stability and aging resistance of polymers.
Article
Polymer Science
J. Gomez-Caturla, J. Ivorra-Martinez, R. Tejada-Oliveros, V Moreno, D. Garcia-Garcia, R. Balart
Summary: This work focuses on the development of environmentally friendly PLA formulations by using different esters derived from geraniol as plasticizers. The results show that these esters have good compatibility with PLA, and they can effectively enhance the elongation and plasticity of PLA, reduce its glass transition temperature, and slightly improve its water absorption capabilities.
Article
Polymer Science
Dandan Li, Zhaohui Lu, Zhao Ke, Ke Xu, Fengna Dai, Youhai Yu, Guangtao Qian, Chunhai Chen
Summary: In this study, cross-linked PI aerogel membranes with low dielectric constant and high moisture resistance were prepared by co-polymerization and scraping coating technology. The incorporation of fluorinated blocks and benzimidazole ring structures resulted in novel PI aerogel membranes with fascinating dielectric properties and outstanding moisture resistance.
Article
Polymer Science
Zhaoyang Chu, Qing Zhang, Haihua Luo, Han Zhou, Fapei Zhang, Wei Chen, Wenhua Zhang
Summary: Multi-physical fields solution processing strategy is a universal and facile method for preparing various conjugated polymer films and high-performance devices. In this study, we employed a combined microfluidic flow and ultrasonication strategy (FU) for processing CP solutions, and found a pronounced synergetic effect in promoting the pre-ordering of chains in solution. The conformation order and anisotropy of the solution were revealed through various characterizations. A non-classical nucleation model for polymer crystallization in non-equilibrium solution processing was confirmed. The roles of microfluidic flow and ultrasonication in chain aggregation and crystallization were addressed through multi-physical simulations. Compared to pristine solutions, the FU strategy showed improved solution anisotropy and crystallization kinetics, resulting in higher crystallinity in films and increased mobilities in OFET devices. The FU processing strategy provides a universal approach for regulating chain conformation and aggregation in conjugated polymer solutions.