Intelligent Identification for Rock-Mineral Microscopic Images Using Ensemble Machine Learning Algorithms.

It is significant to identify rock-mineral microscopic images in geological engineering. The task of microscopic mineral image identification, which is often conducted in the lab, is tedious and time-consuming. Deep learning and convolutional neural networks (CNNs) provide a method to analyze mineral microscopic images efficiently and smartly. In this research, the transfer learning model of mineral microscopic images is established based on Inception-v3 architecture. The four mineral image features, including K-feldspar (Kf), perthite (Pe), plagioclase (Pl), and quartz (Qz or Q), are extracted using Inception-v3. Based on the features, the machine learning methods, logistic regression (LR), support vector machine (SVM), random forest (RF), k-nearest neighbors (KNN), multilayer perceptron (MLP), and gaussian naive Bayes (GNB), are adopted to establish the identification models. The results are evaluated using 10-fold cross-validation. LR, SVM, and MLP have a significant performance among all the models, with accuracy of about 90.0%. The evaluation result shows LR, SVM, and MLP are the outstanding single models in high-dimensional feature analysis. The three models are also selected as the base models in model stacking. The LR model is also set as the meta classifier in the final prediction. The stacking model can achieve 90.9% accuracy, which is higher than all the single models. The result also shows that model stacking effectively improves model performance.

The Splice Isoforms of the Drosophila Ecdysis Triggering Hormone Receptor Have Developmentally Distinct Roles.

To grow, insects must periodically shed their exoskeletons. This process, called ecdysis, is initiated by the endocrine release of Ecdysis Trigger Hormone (ETH) and has been extensively studied as a model for understanding the hormonal control of behavior. Understanding how ETH regulates ecdysis behavior, however, has been impeded by limited knowledge of the hormone's neuronal targets. An alternatively spliced gene encoding a G-protein-coupled receptor (ETHR) that is activated by ETH has been identified, and several lines of evidence support a role in ecdysis for its A-isoform. The function of a second ETHR isoform (ETHRB) remains unknown. Here we use the recently introduced "Trojan exon" technique to simultaneously mutate the ETHR gene and gain genetic access to the neurons that express its two isoforms. We show that ETHRA and ETHRB are expressed in largely distinct subsets of neurons and that ETHRA- but not ETHRB-expressing neurons are required for ecdysis at all developmental stages. However, both genetic and neuronal manipulations indicate an essential role for ETHRB at pupal and adult, but not larval, ecdysis. We also identify several functionally important subsets of ETHR-expressing neurons including one that coexpresses the peptide Leucokinin and regulates fluid balance to facilitate ecdysis at the pupal stage. The general strategy presented here of using a receptor gene as an entry point for genetic and neuronal manipulations should be useful in establishing patterns of functional connectivity in other hormonally regulated networks.

Revisiting the relationship between environmental and financial performance in Chinese industry.

The debate on the relationship between corporate or industrial environmental performance (EP) and financial performance (FP) has yet to be resolved, and studies need to examine the possible moderating effects on the EP-FP link. We argue that industrial EP has a positive effect on FP and that industrial munificence and resource slack can moderate the EP-FP link. Using a dataset from Chinese industrial firms, we examine the direct effect of industrial EP on FP and the indirect effects of industrial munificence and resource slack on the EP-FP link. Our results show that improving corporate or industrial-level EP significantly influences FP and that slack resources play a significant role on the EP-FP link. However, we found no significant moderating effect of industrial munificence on the link.

The relationship between corporate environmental performance and environmental disclosure: an empirical study in China.

Based on a content analysis of 533 Chinese listed companies, this study examines how corporate environmental performance affects not only the level of detail of a company's environmental disclosures, but also what information is disclosed. The results show that (1) both poor and good performers have more disclosure than the median (i.e., "mixed") performers, which provides empirical evidence to support a nonlinear relationship between corporate environmental performance and environmental disclosure; (2) poor performers disclose more soft information on environmental performance than good performers, and good performers disclose more solid information; and (3) although poor performers increase disclosure after being exposed as environmental violators, they avoid disclosing negative environmental information, such as the violation and the associated penalties. This study provides additional evidence for a nonlinear relationship between environmental performance and disclosure in emerging markets, and suggests environmental disclosure may not be a valid signal to differentiate good performers from poor performers in contemporary China.