基于多元神经网络融合的分布式资源空间文本分类研究

针对实体产业对科技资源的服务需求,以服务效应作为资源文本分类标准,提出一种基于多元神经网络融合的分布式资源空间文本分类模型。设计了包含词嵌入层、卷积层、双向门控循环单元层、注意力机制层和softmax层的多元神经网络通路;在此基础上采用基于需求—效应—资源分类策略,完成了从定性科技资源需求到定量资源服务效应求解,再到定性科技资源输出的映射变换,重点解决了分布式科技资源局部和全局语义特征形式多样、文本长距离依赖特征显著、重要资源信息难以准确识别的问题,进而从分布式科技资源空间中快速准确地获取效应知识,提升实体产业产品研发效率和创新能力;通过万方专利科技资源数据集验证了所提方法的可行性和有效性,为更加全面地挖掘资源文本特征和按需服务实体产业提供了一种新的思路和手段。     

A deep neural network based classifier for brain tumor diagnosis

Classification process plays a key role in diagnosing brain tumors. Earlier research works are intended for identifying brain tumors using different classification techniques. However, the False Alarm Rates (FARs) of existing classification techniques are high. To improve the early-stage brain tumor diagnosis via classification the Weighted Correlation Feature Selection Based Iterative Bayesian Multivariate Deep Neural Learning (WCFS-IBMDNL) technique is proposed in this work. The WCFS-IBMDNL algorithm considers medical dataset for classifying the brain tumor diagnosis at an early stage. At first, the WCFS-IBMDNL technique performs Weighted Correlation-Based Feature Selection (WC-FS) by selecting subsets of medical features that are relevant for classification of brain tumors. After completing the feature selection process, the WCFS-IBMDNL technique uses Iterative Bayesian Multivariate Deep Neural Network (IBMDNN) classifier for reducing the misclassification error rate of brain tumor identification. The WCFS-IBMDNL technique was evaluated in JAVA language using Disease Diagnosis Rate (DDR), Disease Diagnosis Time (DDT), and FAR parameter through the epileptic seizure recognition dataset.     

Taxonomic importance of spore morphology in Thelypteridaceae from Northern Pakistan

Spore morphology of Thelypteridaceae species growing in Malakand Division, Northern Pakistan, was studied using both light microscopy and scanning electron microscopy. The taxa are Christella dentata and Glaphyropteridopsis erubescens in the subfamily Thelypteridoideae, and Phegopteris connectilis, Pseudophegopteris pyrrhorhachis, and Pseudophegopteris levingei in the subfamily Phegopteridoideae. The studied species exhibit differences in spore size, exospore thickness, color, and ornamentation. Spores of the studied species are monolete and medium‐sized, and shape is ellipsoidal in both polar and equatorial views. The average measurement of the polar diameter ranges from 27 μm to 31 μm, whereas in the equatorial direction it varied from 20 μm to 40 μm. The exospore thickness ranges from 1.2 μm to 2.4 μm. Reticulate, laevigate with microgranules, cristate, and coarsely echinate surface ornamentation are observed among the species. Multivariate analysis including unweighted pair group method with arithmetic mean and principal component analysis was used for the grouping and discrimination of species and genera.     

Bioecology of macrobenthic communities in the microtidal monsoonal Kodungallur–Azhikode Estuary,southwest coast of India

A comprehensive study of the macrobenthic communities in the Kodungallur–Azhikode Estuary (KAE) was conducted during 2009–2011 period. A total of 18,846 organisms were collected, with 60% being malacostracans, followed by polychaetes (20%), molluscs (9%) and ‘others’ group (11%). A total of 79 species in 71 genera belonging to 49 families were identified, with 33 spp. being polychaetes, 26 spp. being malacostracans, 11 spp. being molluscs and 9 spp. being in the ‘others’ group. A single species of opportunistic amphipod (Americorophium triaeonyx) comprised more than 62.05% of the total numerical abundance of macrobenthos. The other dominant species were Obelia bidentata, Arcuatula senhousia, Cirolana fluviatilis, Prionospio cirrifera and Capitella sp. Canonical correspondence analysis (CCA) results indicated environmental parameters such as water column salinity, turbidity, sediment Eh, substratum type, chlorophyll‐a concentration, depth and organic matter in sediment were the significant factors influencing the distribution of macrobenthic species in the KAE. The present study provides baseline information for future KAE benthic studies.     

基于多变量时间序列及向量自回归机器学习模型的水驱油藏产量预测方法

提出了一种基于多变量时间序列(MTS)及向量自回归(VAR)机器学习模型的水驱油藏产量预测方法,并进行了实例应用。该方法在井网分析的基础上通过MTS分析对注采井组数据进行优选,并将井组内不同采出井产油量及注入井注水量作为彼此相关的时间序列,通过建立VAR模型从多个时间序列中提取出相互作用规律,挖掘注采井间流量的依赖关系从而进行产量预测。水驱油藏历史生产数据分析结果表明,与数值模拟历史拟合结果相比,机器学习模型产量预测结果具有更高精度,同时不确定性分析提升了预测结果的安全性。通过脉冲响应分析对注入井的采油贡献量进行评价,可为注水开发方案调整提供理论指导。     

Coagulation process in Manchega sheep milk from Spain: A path analysis approach

Characteristics of sheep milk are of great interest for the dairy industry, as almost the totality of production is intended for cheesemaking. However, the existing relationships between these variables are complex. This study assessed composition, hygienic quality, coagulation properties, and curd yield of 1,200 individual Manchega sheep milk samples. The aim was to compare the effect of composition and hygienic quality on coagulation and curdling, and to evaluate the relationship between curd yields and the coagulation process and the effect of other features by using path analysis methodologies. Outcomes proved path analysis to be a useful and effective tool to assess these relationships through direct and indirect paths within the same model. Results showed that the factors that had a direct influence on milk coagulation were lactose concentration, casein content, and initial pH of milk. Contrastingly, somatic cells did not seem to have any effect (direct or indirect) on the coagulation process. Factors that directly affected curd yield were fat content, lactose concentration, casein content, and curd moisture. However, technological parameters showed little effect over curd yield.     

An assessment of the efficiency of Spanish schools: evaluating the influence of the geographical,managerial, and socioeconomic features

The aim of the present paper is to explore the efficiency of Spanish schools while simultaneously considering data envelopment analysis (DEA) and multivariate analysis. Test scores from the Program for International Assessment reports are used as outputs while the resources of each institution are considered as inputs to the analysis. The methodology utilized determines the DEA efficiencies under various input/output combinations and the results are interpreted through the application of factor analysis and property‐fitting techniques. The objective of the study is to identify the strengths and weaknesses of each type of school and the connections with the way in which the efficiency is obtained. In the light of the results, the study concludes that there exist differences related with two of the criteria considered: the type of management of the schools and the parental socioeconomic level of the students. However, no differences appear when the geographical location of the institutions is considered to characterize the entities.