ABSTRACT Objective We denote the four major factors related to the development of type 2 diabetes (T2D) as “diabetes factor” (DF); increased insulin resistance (IR); decreased glucose effectiveness (GE); and the first-and-second-phase of insulin secretion (FPIS, SPIS). The level of hemoglobin (Hb) was found to be related to IR and FPIS, but no-known studies focused on its role in relation to SPIS and GE. In this study, we aim to evaluate the relationships between Hb and all four DFs in the same individual. Subjects and methods We randomly enrolled 24,407 men and 24,889 women between 30 and 59 years old. IR, FPIS, SPIS and GE were measured according to equations published in our previous studies. To compare the slopes between Hb and the four DFs with different units, we converted their units to percent of change per unit of increased Hb. Results Age, HDL-cholesterol and GE were higher in women; BMI, blood pressure, LDL-cholesterol, TG, Hb, FPIS, SPIS and IR were higher in men. After they were converted into percentage, the closeness of their relationships to Hb, from the highest to the lowest, were GE, IR, FPIS and SPIS for women and IR, GE, FPIS and SPIS for men. GE was the only one negatively related to Hb. Conclusions Our data showed that IR, FPIS and SPIS were both positively and, GE negatively, related to the Hb in adult Chinese. For women, GE had the closest association with Hb; for men, it was IR. Both phases of insulin secretion had relatively weaker relationships than IR and GE.

Artificial neural network (ANN) has wide applications such as data processing and classification. However, comparing with other classification methods, ANN needs enormous memory space and training time to build the model. This makes ANN infeasible in practical applications. In this paper, we try to integrate the ideas of human learning mechanism with the existing models of ANN. We propose an incremental neural network construction framework for unsupervised learning. In this framework, a neural network is incrementally constructed by the corresponding subnets with individual instances. First, a subnet maps the relation between inputs and outputs for an observed instance. Then, when combining multiple subnets, the neural network keeps the corresponding abilities to generate the same outputs with the same inputs. This makes the learning process unsupervised and inherent in this framework. In our experiment, Reuters-21578 was used as the dataset to show the effectiveness of the proposed method on text classification. The experimental results showed that our method can effectively classify texts with the best F1-measure of 92.5%. It also showed the learning algorithm can enhance the accuracy effectively and efficiently. This framework also validates scalability in terms of the network size, in which the training and testing times both showed a constant trend. This also validates the feasibility of the method for practical uses.

A chemical study of the EtOAc-soluble ethanolic extract of Monascus kaoliang BCRC 31506 (Eurotiaceae) fermented rice led to the isolation of a new furan-3-one derivative along with 9 known compounds, monascin (2), ankaflavin (3), N-trans-feruloyltyramine (4), vanillic acid (5), 4-acetonyl-3,5-dimethoxy-p-quinol (6), (+)-ß-eudesmol (7), trans-phytol (8), aesculetin dimethyl ether (9) and (-)-matairesinol (10). Structure 1 is proposed for monascuskaoliaone on the basis of spectroscopic evidence (NMR, UV, IR, and HRMS). None of the compounds exhibited significant scavenging activity toward DPPH radical in bioautography or spectroscopic assays.