A survey of randomized algorithms for training neural networks

336 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Randomness in neural networks: an overview (2017, background): “Specifically, the special issue contains an informative editorial by Wang(133) and a survey paper by Zhang and Suganthan.(134) Apart from the algorithms detailed here, the survey covers additional methods that venture outside the definition of randomization we considered, such as probabilistic NNs,(”
Machine Learning and Integrative Analysis of Biomedical Big Data (2019, background): “ANN can be trained in a single step without iterative tuning of hidden node parameters, as opposed to a back-propagation (BP) algorithm which is time-consuming, converges slowly, and can be stuck at local minima [183].”
An Adaptive Multi-Sensor Data Fusion Method Based on Deep Convolutional Neural Networks for Fault Diagnosis of Planetary Gearbox (2017, background): “Although there are several researches [39,40] about the automatic optimization of parameters of DCNN, the optimized process is usually time-consuming and easily converges into a local optimum due to the large number of parameters of DCNN.”
Empirical Mode Decomposition based ensemble deep learning for load demand time series forecasting (2017, methodology): “…i and the neuron j of the hidden layer; w jo is the weight of the between the hidden layer neuron j and the output; b i and b j are the To train an SLFN, random values are assigned , then the weights are tuned by certain method such as back-propagation [37] or using a closed form solution [38, 39].”