References in periodicals archive ?
The initial value of fitness value of BPNN algorithm is about 0.23, and then it rises in the tortuosity.
And the multi-objective optimization methodology was proposed to efficiently optimize the process parameters, which combines LHS, Kriging, BPNN, NSGA-II, and Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) .
For the BPNN models, the optimal number of neurons in the hidden layer was calculated by using leave-one-out cross validation in the range from 10 to 50 (large number of neurons in the hidden layer had no performance improvement in the experiment).
A brief explanation of the BPNN, CpNN, and CNN is given.
In addition, the findings indicated that the SVM models provide better (or at least comparable) performance than BPNN models and do not over-fit the data.
In , a BPNN used cross entropy for calculating the performance of a network with targets and outputs.
Then, according to the difference of the time domain resistance curve for various molecules, the GFCS was endued with selectivity after PCA and BPNN.
Backpropagation neural network (BPNN) which will be used in this research can be considered the most fundamental and widely used method among the ANN methods.
The learning process is divided into two parts: the data collection stage and the BPNN training stage.
The system used a back-propagation neural network (BPNN) to recognize the weed seed image with 94% accuracy.
A comparison was made with a neural network with standard back-propagation (BPNN) algorithm (Hagan and Menhaj, 1994), and similar architecture and database.
 formulated a correlation model with the backpropagation neural network (BPNN) technique to eliminate the temperature effect.
Acronyms browser ?
Full browser ?