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
This study proposed a new royal crown-shaped polarisation insensitive double negative triple band microwave range electromagnetic metamaterial absorber(MA).The primary purpose of this study is to utilise the exotic ch...This study proposed a new royal crown-shaped polarisation insensitive double negative triple band microwave range electromagnetic metamaterial absorber(MA).The primary purpose of this study is to utilise the exotic characteristics of this perfect metamaterial absorber(PMA)for microwave wireless communications.The fundamental unit cell of the proposed MA consists of two pentagonal-shaped resonators and two inverse C-shaped metallic components surrounded by a split ring resonator(SRR).The bottom thin copper deposit and upper metallic resonator surface are disjoined by an FR-4 dielectric substrate with 1.6 mm thickness.The CST MW studio,a high-frequency electromagnetic simulator has been deployed for numerical simula-tion of the unit cell in the frequency range of 4 to 14 GHz.In the TE mode,the offered MA structure demonstrated three different absorption peaks at 6.85 GHz(C-band),8.87 GHz(X-band),and 12.03 GHz(Ku-band),with 96.82%,99.24%,and 99.43%absorptivity,respectively.The electric field,magnetic field,and surface current distribution were analysed using Maxwell’s-Curl equations,whereas the angle sensitivity was investigated to comprehend the absorption mechanism of the proposed absorber.The numerical results were verified using the Ansys HFSS(high-frequency structure simulator)and ADS(advanced design system)for equivalent circuit models.Moreover,the proposed MA is polarisation and incident angle independent.Hence,the application of this MA can be extended to a great extent,including airborne radar applications,defence,and stealth-coating technology.展开更多
The CuxO/TiO_(2)nanotubes arrays are fabricated in two stages.Firstly,TiO_(2)-NTs are grown by the Ti-foil anodization process and then annealed for 2h at 500℃.Subsequently,CuxO thin film was deposited with different...The CuxO/TiO_(2)nanotubes arrays are fabricated in two stages.Firstly,TiO_(2)-NTs are grown by the Ti-foil anodization process and then annealed for 2h at 500℃.Subsequently,CuxO thin film was deposited with different deposition times on the nanotubes by electrochemical cathodic reaction,then heated twice,once at 200℃in the air and then at 300℃ in the closed furnace for 2 h,respectively.Pure-TNT and Cu_(x)O/TNTs heterostructure are characterized by XRD,FE-SEM,EDX,Hall effect,and as a gas sensor.Results show that the gas sensor(CuO_(x=1)/TiO_(2)for NO_(2)and H_(2)gases)prepared at the time(1 min)is higher than the pure TiO_(2)-NTs and also higher than Cu_(x=2)O/TiO_(2)which were synthesized at various times 3,5,7,and 10 mins.展开更多
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.
基金supported by Fundamental Research Grant Scheme(FRGS),MOE,Malaysia,Code:FRGS/1/2022/TK07/UKM/02/22.
文摘This study proposed a new royal crown-shaped polarisation insensitive double negative triple band microwave range electromagnetic metamaterial absorber(MA).The primary purpose of this study is to utilise the exotic characteristics of this perfect metamaterial absorber(PMA)for microwave wireless communications.The fundamental unit cell of the proposed MA consists of two pentagonal-shaped resonators and two inverse C-shaped metallic components surrounded by a split ring resonator(SRR).The bottom thin copper deposit and upper metallic resonator surface are disjoined by an FR-4 dielectric substrate with 1.6 mm thickness.The CST MW studio,a high-frequency electromagnetic simulator has been deployed for numerical simula-tion of the unit cell in the frequency range of 4 to 14 GHz.In the TE mode,the offered MA structure demonstrated three different absorption peaks at 6.85 GHz(C-band),8.87 GHz(X-band),and 12.03 GHz(Ku-band),with 96.82%,99.24%,and 99.43%absorptivity,respectively.The electric field,magnetic field,and surface current distribution were analysed using Maxwell’s-Curl equations,whereas the angle sensitivity was investigated to comprehend the absorption mechanism of the proposed absorber.The numerical results were verified using the Ansys HFSS(high-frequency structure simulator)and ADS(advanced design system)for equivalent circuit models.Moreover,the proposed MA is polarisation and incident angle independent.Hence,the application of this MA can be extended to a great extent,including airborne radar applications,defence,and stealth-coating technology.
基金supported by the University of Al-Qadisiyah,Iraq,and Imam Abdulrahman bin Faisal University,Saudi Arabia.
文摘The CuxO/TiO_(2)nanotubes arrays are fabricated in two stages.Firstly,TiO_(2)-NTs are grown by the Ti-foil anodization process and then annealed for 2h at 500℃.Subsequently,CuxO thin film was deposited with different deposition times on the nanotubes by electrochemical cathodic reaction,then heated twice,once at 200℃in the air and then at 300℃ in the closed furnace for 2 h,respectively.Pure-TNT and Cu_(x)O/TNTs heterostructure are characterized by XRD,FE-SEM,EDX,Hall effect,and as a gas sensor.Results show that the gas sensor(CuO_(x=1)/TiO_(2)for NO_(2)and H_(2)gases)prepared at the time(1 min)is higher than the pure TiO_(2)-NTs and also higher than Cu_(x=2)O/TiO_(2)which were synthesized at various times 3,5,7,and 10 mins.
