RESUMEN
While supervised learning of over-parameterized neural networks achieved state-of-the-art performance in image classification, it tends to over-fit the labeled training samples to give inferior generalization ability. Output regularization deals with over-fitting by using soft targets as additional training signals. Although clustering is one of the most fundamental data analysis tools for discovering general-purpose and data-driven structures, it has been ignored in existing output regularization approaches. In this article, we leverage this underlying structural information by proposing Cluster-based soft targets for Output Regularization (CluOReg). This approach provides a unified way for simultaneous clustering in embedding space and neural classifier training with cluster-based soft targets via output regularization. By explicitly calculating a class relationship matrix in the cluster space, we obtain classwise soft targets shared by all samples in each class. Results of image classification experiments under various settings on a number of benchmark datasets are provided. Without resorting to external models or designed data augmentation, we get consistent and significant reductions in classification error compared with other approaches, demonstrating that cluster-based soft targets effectively complement the ground-truth label.
RESUMEN
The use of hemoglobin (Hb) to drive atom transfer radical polymerization (ATRP) process (Hb-ATRP) for detection of lung cancer related nucleic acid is firstly reported. Hb does not need to be treated prior to using indicating the potential for synthetic engineering in complex biological microenvironments without the need for in vitro techniques. Here, we report a new signal amplification strategy using Hb-mediated graft of nitronyl niroxide monoradical polymers as a signal-on electrochemical biosensor for ultralow level DNA highly selective detection. Building DNA biosensors includes: (i) the fixation of peptide nucleic acid (PNA) probe (no phosphate group) via the 5' terminus-SH; (ii) the modification of transition metal; (iii) Site-specific markers of Hb-ATRP promoter, and (iv) the grafting of polymers with electrochemical signal by Hb-ATRP process. Through the Hb-ATRP process of nitronyl nitroxide monoradical (TEMPO), the presence of a small amount of DNA can eventually result in calling a certain number of TEMPO redox tags. Obviously, the Hb-ATRP is a method of easy source of raw materials, simple operation and no need for complex equipment. The constructed biosensor, as expected, is highly selective and sensitive to target DNA. The detection limit can be calculated as 15.96 fM under optimal conditions. The excellent performance also shows that the constructed DNA biosensor is suitable for DNA screening and DNA concentration determination in complex sample matrix.
