Visual diagnosis of skin cancer is challenging due to subtle inter-class similarities,variations in skin texture,the presence of hair,and inconsistent illumination.Deep learning models have shown promise in assisting ...Visual diagnosis of skin cancer is challenging due to subtle inter-class similarities,variations in skin texture,the presence of hair,and inconsistent illumination.Deep learning models have shown promise in assisting early detection,yet their performance is often limited by the severe class imbalance present in dermoscopic datasets.This paper proposes CANNSkin,a skin cancer classification framework that integrates a convolutional autoencoder with latent-space oversampling to address this imbalance.The autoencoder is trained to reconstruct lesion images,and its latent embeddings are used as features for classification.To enhance minority-class representation,the Synthetic Minority Oversampling Technique(SMOTE)is applied directly to the latent vectors before classifier training.The encoder and classifier are first trained independently and later fine-tuned end-to-end.On the HAM10000 dataset,CANNSkin achieves an accuracy of 93.01%,a macro-F1 of 88.54%,and an ROC–AUC of 98.44%,demonstrating strong robustness across ten test subsets.Evaluation on the more complex ISIC 2019 dataset further confirms the model’s effectiveness,where CANNSkin achieves 94.27%accuracy,93.95%precision,94.09%recall,and 99.02%F1-score,supported by high reconstruction fidelity(PSNR 35.03 dB,SSIM 0.86).These results demonstrate the effectiveness of our proposed latent-space balancing and fine-tuned representation learning as a new benchmark method for robust and accurate skin cancer classification across heterogeneous datasets.展开更多
Credit risk assessment is an important task of risk management for financial institutions.Machine learning-based approaches have made promising progress in credit risk assessment by treating it as imbalanced binary cl...Credit risk assessment is an important task of risk management for financial institutions.Machine learning-based approaches have made promising progress in credit risk assessment by treating it as imbalanced binary classification tasks.However,few efforts have been made to deal with the class overlap problem that accompanies imbalances simultaneously.To this end,this study proposes a Tomek link and genetic algorithm(GA)-based under-sampling framework(TEUS)to address the class imbalance and overlap issues in binary credit classification by eliminating majority class instances with considering multi-perspective factors.TEUS first determines boundary majority instances with Tomek link,then take the distance from each majority instance to its nearest boundary as the radius and assigns the density of opposite class samples within the radius as the overlap potential of that majority instance.Second,TEUS weighs each non-borderline majority instance based on its information contribution in estimating class labels.After partitioning non-borderline majority instances into subgroups according to overlap potential and information contribution,TEUS applies GA to select samples from subgroups and merge them with the minority samples into a new training set.Innovatively,the design of the fitness function in GA and the grouping of the non-borderline majority not only trade off the multi-perspective characteristics of instances but also help reduce the computational complexity of the sampling optimization search.Numerical experiments on real-world credit data sets demonstrate the effectiveness of the proposed TEUS.展开更多
基金supported and funded by the Deanship of Scientific Research at Imam Mohammad Ibn Saud Islamic University(IMSIU)(grant number IMSIU-DDRSP2601).
文摘Visual diagnosis of skin cancer is challenging due to subtle inter-class similarities,variations in skin texture,the presence of hair,and inconsistent illumination.Deep learning models have shown promise in assisting early detection,yet their performance is often limited by the severe class imbalance present in dermoscopic datasets.This paper proposes CANNSkin,a skin cancer classification framework that integrates a convolutional autoencoder with latent-space oversampling to address this imbalance.The autoencoder is trained to reconstruct lesion images,and its latent embeddings are used as features for classification.To enhance minority-class representation,the Synthetic Minority Oversampling Technique(SMOTE)is applied directly to the latent vectors before classifier training.The encoder and classifier are first trained independently and later fine-tuned end-to-end.On the HAM10000 dataset,CANNSkin achieves an accuracy of 93.01%,a macro-F1 of 88.54%,and an ROC–AUC of 98.44%,demonstrating strong robustness across ten test subsets.Evaluation on the more complex ISIC 2019 dataset further confirms the model’s effectiveness,where CANNSkin achieves 94.27%accuracy,93.95%precision,94.09%recall,and 99.02%F1-score,supported by high reconstruction fidelity(PSNR 35.03 dB,SSIM 0.86).These results demonstrate the effectiveness of our proposed latent-space balancing and fine-tuned representation learning as a new benchmark method for robust and accurate skin cancer classification across heterogeneous datasets.
基金supported by National Key R&D Programof ChinaunderGrant No.2019YFB1404600Beijing Natural Science Funds under Grant No.9162003Beijing's"High-grade,Precision and Advanced Discipline Construction(Municipal)-Business Administration"project under Grant No.19008022065.
文摘Credit risk assessment is an important task of risk management for financial institutions.Machine learning-based approaches have made promising progress in credit risk assessment by treating it as imbalanced binary classification tasks.However,few efforts have been made to deal with the class overlap problem that accompanies imbalances simultaneously.To this end,this study proposes a Tomek link and genetic algorithm(GA)-based under-sampling framework(TEUS)to address the class imbalance and overlap issues in binary credit classification by eliminating majority class instances with considering multi-perspective factors.TEUS first determines boundary majority instances with Tomek link,then take the distance from each majority instance to its nearest boundary as the radius and assigns the density of opposite class samples within the radius as the overlap potential of that majority instance.Second,TEUS weighs each non-borderline majority instance based on its information contribution in estimating class labels.After partitioning non-borderline majority instances into subgroups according to overlap potential and information contribution,TEUS applies GA to select samples from subgroups and merge them with the minority samples into a new training set.Innovatively,the design of the fitness function in GA and the grouping of the non-borderline majority not only trade off the multi-perspective characteristics of instances but also help reduce the computational complexity of the sampling optimization search.Numerical experiments on real-world credit data sets demonstrate the effectiveness of the proposed TEUS.
