A deep learning-based method for grip strength prediction: Comparison of multilayer perceptron and polynomial regression approaches

Hwang

Journal of Occupational Health

2021

Self Cite

Section: Discussionsupporting

confidence: 82%

Prediction and comparison of postural discomfort based on MLP and quadratic regression

Hwang

Journal of Occupational Health

2021

Self Cite

“…The main difference with a multiple regression model is that the weights on the neural network cannot be interpreted as in the multiple regression model, but the magnitude of the weights only indicates the strength or weakness of the relationship between two adjacent nodes. In addition, in the neural network model, each node uses a certain formula called the activation function which is generally a non-linear function [42]. The diagram of an MLP neural network model is illustrated in Fig.…”

Section: Multi-layer Perceptron Neural Networkmentioning

confidence: 99%

Modeling Multioutput Response Uses Ridge Regression and MLP Neural Network with Tuning Hyperparameter through Cross Validation

Nugroho¹,

Handoyo²,

Hsieh³

et al. 2022

IJACSA

The multiple regression model is very popular among researchers in both field of social and science because it is easy to interpret and have a well-established theoretical framework. However, the multioutput multiple regression model is actually widely applied in the engineering field because in the industrial world there are many systems with multiple outputs. The ridge regression model and the Multi-Layer Perceptron (MLP) neural network model are representations of the predictive linear regression model and predictive non-linear regression model that are widely applied in the world of practice. This study aims to build multi-output models of a ridge regression model and an MLP neural network whose hyperparameters are determined by a grid search algorithm through the cross-validation method. The hyperparameter that produces the smallest RMSE value in the validation data is chosen as the hyperparameter to train both models on the training data. The hyperparameter in question is a combination of learning algorithms and alpha values (ridge regression), a combination of the number of hidden nodes and gamma values (MLP neural network). In the ridge regression model for alpha in the range between 0.1 and 0.7, the smallest RMSE is obtained for all learning algorithms used. While the MLP neural network model specifically obtained a combination of the number of nodes = 18 and gamma = 0.1 which produces the smallest RMSE. The ridge regression model with selected hyperparameters has better performance (in the RMSE and R2 value) than the MLP neural network model with selected hyperparameters, both on training and testing data.

Computational Intelligence and Neuroscience

“…The artificial neuron receives the given signal of the previous neuron, and each given signal will attach a weight. Under the combined action of all weights, this neuron will show a corresponding activation state [ 10 – 12 ], expressed as follows:

where f ( x ) is the final output, x i is the input signal, and w i represents the weight corresponding to the input signal, with n groups in total.…”

Section: Construction Of the Scoring System For English Essaysmentioning

confidence: 99%

Machine Learning-Based Intelligent Scoring System for English Essays under the Background of Modern Information Technology

Hong-fu

2022

This work is to reduce the workload of teachers in English teaching and improve the writing level of students, so as to provide a way for students to practice English composition scoring independently and satisfy the needs of college teachers and students for intelligent English composition scoring and intelligently generated comments. In this work, it firstly clarifies the teaching requirements of college English classrooms and expounds the principles and advantages of machine learning technology. Secondly, a three-layer neural network model (NNM) is constructed by using the multilayer perceptron (MLP), combined with the latent Dirichlet allocation (LDA) algorithm. Furthermore, three semantic representation vector technologies, including word vector, paragraph vector, and full-text vector feature, are used to represent the full-text vocabulary of English composition. Then, a model based on the K-nearest neighbors (kNN) algorithm is proposed to generate English composition evaluation, and a final score based on the extreme gradient boosting (XGBoost) model is proposed. Finally, a model dataset is constructed using 800 college students’ English essays for the CET-4 mock test, and the model is tested. The research results show that the semantic representation vector technology proposed can more effectively extract the lexical semantic features of English compositions. The XGBoost model and the kNN algorithm model are used to score and evaluate English compositions, which improves the accuracy of the scores. This makes the management of the entire scoring model more efficient and more accurate. It means that the model proposed is better than the traditional model in terms of evaluation accuracy. This work provides a new direction for the application of artificial intelligence technology in English teaching under the background of modern information technology.