Article
Computer Science, Artificial Intelligence
Ruiyuan Lin, Zhiruo Zhou, Suya You, Raghuveer Rao, C. -C. Jay Kuo
Summary: This work interprets the multilayer perceptron (MLP) neural network from a geometrical viewpoint and proposes a new three-layer feedforward MLP (FF-MLP) architecture for its implementation. Experiments show that the FF-MLP outperforms the traditional backpropagation-based MLP (BP-MLP) in terms of design time, training time, and classification performance.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Geochemistry & Geophysics
Guangsheng Chen, Hailiang Lu, Donglin Di, Linhui Li, Mahmoud Emam, Weipeng Jing
Summary: In this study, a deep-learning-based method called spatiotemporal fusion multilayer perceptron (StfMLP) is proposed to achieve more accurate remote-sensing image fusion with a small-scale of data. Experimental results demonstrate that the proposed method outperforms the state-of-the-art methods effectively on two public datasets.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Health Care Sciences & Services
Rahman Ali, Jamil Hussain, Seung Won Lee
Summary: In this study, a feed-forward artificial neural network (ANN)-based self-care prediction methodology, called multilayer perceptron (MLP)-progressive, has been proposed to improve the early detection of self-care disabilities in children. The proposed MLP-progressive model outperforms existing methods and achieves a classification accuracy of 97.14% and 98.57% on multi-class and binary-class datasets, respectively.
Article
Computer Science, Artificial Intelligence
Xuetao Xie, Yi-Fei Pu, Jian Wang
Summary: This paper proposes a fractional gradient descent (RFGD) algorithm that demonstrates strong robustness to the initial weights of multilayer perceptron (MLP). Experimental results show that the RFGD algorithm outperforms other algorithms in terms of robustness for the initial weights of MLP. The effectiveness and convergence of the algorithm are also analyzed.
Article
Energy & Fuels
Mohamed Trabelsi, Mohamed Massaoudi, Ines Chihi, Lilia Sidhom, Shady S. Refaat, Tingwen Huang, Fakhreddine S. Oueslati
Summary: This paper proposes a hybrid model combining SR and MLP for PV power forecasting. The model has advantages in accuracy and robustness by eliminating unimportant inputs, preserving hyperparameters, using fewer layers and neurons, and reducing the number of iterations.
Article
Computer Science, Artificial Intelligence
Odai Y. Dweekat, Sarah S. Lam
Summary: Diabetes is a challenging and threatening disease, and data mining techniques have been applied to predict and classify diabetic patients. This research integrates design of experiments (DOE), genetic algorithm (GA), and multilayer perceptron (MLP) to classify diabetic patients. The proposed approach outperforms eight different classification algorithms and presents a robust predictive tool for early detection of diabetes.
Article
Computer Science, Artificial Intelligence
Qinghua Jiang, Lailai Zhu, Chang Shu, Vinothkumar Sekar
Summary: This work introduces an efficient multilayer RBF network by combining MLPs and RBF-NNs, achieving better approximation accuracy and faster training convergence for regression problems through nonlinear transformation of input data using multivariate basis functions.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Health Care Sciences & Services
Odai Y. Dweekat, Sarah S. Lam
Summary: This research proposes an integrated system of Genetic Algorithm, Multilayer Perceptron, and Principal Component Analysis that accurately predicts cervical cancer. By optimizing hyperparameters and transforming features, the proposed method outperforms other classification algorithms and can be used for early prediction of cervical cancer.
Article
Engineering, Multidisciplinary
Vijay Kumar Singh, Kanhu Charan Panda, Atish Sagar, Nadhir Al-Ansari, Huan-Feng Duan, Pradosh Kumar Paramaguru, Dinesh Kumar Vishwakarma, Ashish Kumar, Devendra Kumar, P. S. Kashyap, R. M. Singh, Ahmed Elbeltagi
Summary: Two hybrid Machine Learning based PTFs combining Genetic Algorithm with Multilayer Perceptron and Support Vector Machine were proposed to estimate soil hydraulic conductivity, with SVM outperforming other methods. The new model facilitates efficient measurement of hydraulic conductivity using pre-available databases.
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS
(2022)
Article
Engineering, Biomedical
Muhammad Ibrahim Dutt, Wala Saadeh
Summary: This paper proposes a robust and computationally efficient framework for predicting and evaluating the depth of anesthesia in patients. The framework utilizes a deep learning model that incorporates wavelet transform and fractal features to achieve accurate estimation, regardless of age and anesthetic agent type.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2023)
Article
Computer Science, Interdisciplinary Applications
Hong Wang, Hossein Moayedi, Loke Kok Foong
Summary: The study aimed to optimize a design combining artificial neural network (ANN) and a genetic algorithm (GA) for calculating slope stability safety factors in pure cohesive slopes. Through a series of trial and error processes, the optimized model showed proper performance with the least mean square error, utilizing a multi-layer perceptron algorithm. The optimal architecture and statistical indices of the model were evaluated for accuracy and performance.
ENGINEERING WITH COMPUTERS
(2021)
Article
Energy & Fuels
Hui Wang, Jilong Wang
Summary: The vigorous development of wind energy is crucial for adjusting energy structure, controlling atmospheric smog, and transforming economic development mode. However, the random and volatile nature of wind speed makes accurate prediction challenging. In this study, the Multi-Task Lasso method is applied for variable selection to improve the accuracy of short-term wind speed prediction. The extracted features are then used by the Multilayer Perceptron for prediction modeling. Experimental results demonstrate the significant impact of feature selection on prediction accuracy, achieving an improvement of over 17%.
Article
Engineering, Environmental
Roya Narimani, Changhyun Jun, Carlo De Michele, Thian Yew Gan, Somayeh Moghimi Nezhad, Jongyun Byun
Summary: This study proposes a novel approach using Multilayer Perceptron (MLP) neural networks to estimate missing rainfall data. The approach considers three configurations representing different seasons and variations. The rainfall dataset was transformed using the wavelet transform method and a mathematical model was created to analyze and predict the transformed data. Missing rainfall data in Seoul station were reconstructed using the transformed data from other stations. Results showed that the Coi_MLP model with Coiflet wavelet transform accurately estimated missing data.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2023)
Article
Engineering, Civil
Duc Dam Nguyen, Panayiotis C. Roussis, Binh Thai Pham, Maria Ferentinou, Anna Mamou, Dung Quang Vu, Quynh-Anh Thi Bui, Duong Kien Trong, Panagiotis G. Asteris
Summary: This research developed machine learning models to predict the swelling potential of fine-grained soils based on seasonal moisture variations. The results showed that the Bagging-MLP model performed with the highest prediction accuracy, demonstrating a promising data-centric approach for supporting geotechnical design.
TRANSPORTATION GEOTECHNICS
(2022)
Article
Energy & Fuels
Paulino Jose Garcia Nieto, Esperanza Garcia-Gonzalo, Beatriz M. Paredes-Sanchez, Jose P. Paredes-Sanchez
Summary: This study developed an artificial smart model based on support vector machines and grid search optimizer for predicting and characterizing the Higher Heating Value (HHV) of raw biomass. The results showed that the model was accurate in predicting the HHV of biomass and highlighted the importance of physico-chemical parameters in determining the HHV.