Radial Basis Function Neural Network(RBFNN)ensembles have long suffered from non-efficient training,where incorrect parameter settings can be computationally disastrous.This paper examines different evolutionary algor...Radial Basis Function Neural Network(RBFNN)ensembles have long suffered from non-efficient training,where incorrect parameter settings can be computationally disastrous.This paper examines different evolutionary algorithms for training the Symbolic Radial Basis Function Neural Network(SRBFNN)through the behavior’s integration of satisfiability programming.Inspired by evolutionary algorithms,which can iteratively find the nearoptimal solution,different Evolutionary Algorithms(EAs)were designed to optimize the producer output weight of the SRBFNN that corresponds to the embedded logic programming 2Satisfiability representation(SRBFNN-2SAT).The SRBFNN’s objective function that corresponds to Satisfiability logic programming can be minimized by different algorithms,including Genetic Algorithm(GA),Evolution Strategy Algorithm(ES),Differential Evolution Algorithm(DE),and Evolutionary Programming Algorithm(EP).Each of these methods is presented in the steps in the flowchart form which can be used for its straightforward implementation in any programming language.With the use of SRBFNN-2SAT,a training method based on these algorithms has been presented,then training has been compared among algorithms,which were applied in Microsoft Visual C++software using multiple metrics of performance,including Mean Absolute Relative Error(MARE),Root Mean Square Error(RMSE),Mean Absolute Percentage Error(MAPE),Mean Bias Error(MBE),Systematic Error(SD),Schwarz Bayesian Criterion(SBC),and Central Process Unit time(CPU time).Based on the results,the EP algorithm achieved a higher training rate and simple structure compared with the rest of the algorithms.It has been confirmed that the EP algorithm is quite effective in training and obtaining the best output weight,accompanied by the slightest iteration error,which minimizes the objective function of SRBFNN-2SAT.展开更多
The study of training hyperparameters optimisation problems remains underexplored in skin lesion research.This is the first report of using hierarchical optimisation to improve computational effort in a four-dimension...The study of training hyperparameters optimisation problems remains underexplored in skin lesion research.This is the first report of using hierarchical optimisation to improve computational effort in a four-dimensional search space for the problem.The authors explore training parameters selection in optimising the learning process of a model to differentiate pigmented lesions characteristics.In the authors'demonstration,pretrained GoogleNet is fine-tuned with a full training set by varying hyperparameters,namely epoch,mini-batch value,initial learning rate,and gradient threshold.The iterative search of the optimal global-local solution is by using the derivative-based method.The authors used non-parametric one-way ANOVA to test whether the classification accuracies differed for the variation in the training parameters.The authors identified the mini-batch size and initial learning rate as parameters that significantly influence the model's learning capability.The authors'results showed that a small fraction of combinations(5%)from constrained global search space,in contrarily to 82%at the local level,can converge with early stopping conditions.The mean(standard deviation,SD)validation accuracies increased from 78.4(4.44)%to 82.9(1.8)%using the authors'system.The fine-tuned model's performance measures evaluated on a testing dataset showed classification accuracy,precision,sensitivity,and specificity of 85.3%,75.6%,64.4%,and 97.2%,respectively.The authors'system achieves an overall better diagnosis performance than four state-of-the-art approaches via an improved search of parameters for a good adaptation of the model to the authors'dataset.The extended experiments also showed its superior performance consistency across different deep networks,where the overall classification accuracy increased by 5%with this technique.This approach reduces the risk of search being trapped in a suboptimal solution,and its use may be expanded to network architecture optimisation for enhanced diagnostic performance.展开更多
基金This work is supported by Ministry of Higher Education(MOHE)through Fundamental Research Grant Scheme(FRGS)(FRGS/1/2020/STG06/UTHM/03/7).
文摘Radial Basis Function Neural Network(RBFNN)ensembles have long suffered from non-efficient training,where incorrect parameter settings can be computationally disastrous.This paper examines different evolutionary algorithms for training the Symbolic Radial Basis Function Neural Network(SRBFNN)through the behavior’s integration of satisfiability programming.Inspired by evolutionary algorithms,which can iteratively find the nearoptimal solution,different Evolutionary Algorithms(EAs)were designed to optimize the producer output weight of the SRBFNN that corresponds to the embedded logic programming 2Satisfiability representation(SRBFNN-2SAT).The SRBFNN’s objective function that corresponds to Satisfiability logic programming can be minimized by different algorithms,including Genetic Algorithm(GA),Evolution Strategy Algorithm(ES),Differential Evolution Algorithm(DE),and Evolutionary Programming Algorithm(EP).Each of these methods is presented in the steps in the flowchart form which can be used for its straightforward implementation in any programming language.With the use of SRBFNN-2SAT,a training method based on these algorithms has been presented,then training has been compared among algorithms,which were applied in Microsoft Visual C++software using multiple metrics of performance,including Mean Absolute Relative Error(MARE),Root Mean Square Error(RMSE),Mean Absolute Percentage Error(MAPE),Mean Bias Error(MBE),Systematic Error(SD),Schwarz Bayesian Criterion(SBC),and Central Process Unit time(CPU time).Based on the results,the EP algorithm achieved a higher training rate and simple structure compared with the rest of the algorithms.It has been confirmed that the EP algorithm is quite effective in training and obtaining the best output weight,accompanied by the slightest iteration error,which minimizes the objective function of SRBFNN-2SAT.
基金Number:FRGS/1/2020/TK0/UTHM/02/27Universiti Tun Hussein Onn Malaysia,Grant/Award Number:H766。
文摘The study of training hyperparameters optimisation problems remains underexplored in skin lesion research.This is the first report of using hierarchical optimisation to improve computational effort in a four-dimensional search space for the problem.The authors explore training parameters selection in optimising the learning process of a model to differentiate pigmented lesions characteristics.In the authors'demonstration,pretrained GoogleNet is fine-tuned with a full training set by varying hyperparameters,namely epoch,mini-batch value,initial learning rate,and gradient threshold.The iterative search of the optimal global-local solution is by using the derivative-based method.The authors used non-parametric one-way ANOVA to test whether the classification accuracies differed for the variation in the training parameters.The authors identified the mini-batch size and initial learning rate as parameters that significantly influence the model's learning capability.The authors'results showed that a small fraction of combinations(5%)from constrained global search space,in contrarily to 82%at the local level,can converge with early stopping conditions.The mean(standard deviation,SD)validation accuracies increased from 78.4(4.44)%to 82.9(1.8)%using the authors'system.The fine-tuned model's performance measures evaluated on a testing dataset showed classification accuracy,precision,sensitivity,and specificity of 85.3%,75.6%,64.4%,and 97.2%,respectively.The authors'system achieves an overall better diagnosis performance than four state-of-the-art approaches via an improved search of parameters for a good adaptation of the model to the authors'dataset.The extended experiments also showed its superior performance consistency across different deep networks,where the overall classification accuracy increased by 5%with this technique.This approach reduces the risk of search being trapped in a suboptimal solution,and its use may be expanded to network architecture optimisation for enhanced diagnostic performance.
