电网故障处置预案文本中的命名实体识别研究

电网故障处置预案是电网故障处置的重要参考,对电网故障处置预案文本中各类电力设备、名称编号等细粒度的关键实体信息进行抽取,是实现计算机学习理解预案内容并进一步支撑故障处置智能化的重要基础。文中提出一种基于深度学习的电网故障处置预案文本命名实体识别技术,首先采用字向量表征预案文本,然后将注意力机制以及双向长短期记忆网络相结合,有所侧重地提取实体词深层字符特征,最后采用条件随机场求解最优序列化的标注。算例表明:文中所提预案文本命名实体识别模型不依赖人工特征,能够自动高效地提取文本特征,准确识别预案文本中细粒度的实体词,满足预案文本中关键实体信息精确定位和识别的要求。     

基于图像识别的铜转炉吹炼造渣期渣含Fe预测模型研究

铜转炉吹炼是火法炼铜的关键工序,其终点判断与炉寿、铜产率和直收率紧密相关,目前现有人工经验、仪器测定和物料平衡法等终点判断方法均存在一定的局限性。理论上铜转炉吹炼造渣期终点与渣含Fe是否达标有关,而不同Fe含量渣样呈现不同的图像特征,鉴于此,基于图形识别的特征向量提取原理,分别采用卷积神经网络(CNN)算法与支持向量机(SVM)算法,构建了铜转炉吹炼造渣期渣含Fe预测模型,为图像识别技术在铜转炉吹炼终点判断中的应用奠定数模基础。两种模型的实例分析表明,卷积神经网络的训练集预测准确率98%,测试集预测准确率约50%;支持向量机模型的训练集预测准确率99%,测试集预测准确率62%。     

Quadratic optimization fine tuning for the Support Vector Machines learning phase

This work presents a comparative analysis of specific, rather than general, mathematical programming implementation techniques of the quadratic optimization problem (QP) based on Support Vector Machines (SVM) learning process. Considering the Karush–Kuhn–Tucker (KKT) optimality conditions, we present a strategy of implementation of the SVM-QP following three classical approaches: (i) active set, also divided in primal and dual spaces, methods, (ii) interior point methods and (iii) linearization strategies. We also present the general extension to treat large-scale applications consisting in a general decomposition of the QP problem into smaller ones, conserving the exact solution approach. In the same manner, we propose a set of heuristics to take into account for a better than a random selection process for the initialization of the decomposition strategy. We compare the performances of the optimization strategies using some well-known benchmark databases.     

基于运动相关皮层电位握力运动模式识别研究

面向基于脑-机接口（Brain-computer interface,BCI）的脑-机交互控制（Brain-machine interaction control,BMIC）——直接脑控机器人,提出一种新的左、右手握力运动参数范式,在该范式下探索左、右手握力运动相关皮层电位/运动相关电位（Movement-related potentials,MRPs）的时域特征表示并识别握力运动模式.在涉及左、右手4个不同任务的实验中采集了11个健康被试的脑电信号,任务期间要求被试以2种握力变化模式之一完成自愿握力运动,每种任务随机重复30次.不同握力任务之间具有显著差异的运动相关电位特征用于识别握力运动模式.分别用基于核的Fisher线性判别分析和支持向量机识别4个不同的握力运动任务.研究结果进一步证实运动相关电位可以表征握力运动规划、运动执行和运动监控的脑神经机制过程.基于核的Fisher线性判别分析和支持向量机分别获得24±4%和21±5%的平均错误分类率.最小误分类率是12%,所有被试平均最小误分类率为20.9±5%.与传统的仅仅识别参与运动的肢体类型以及识别单侧肢体运动参数的研究相比,本研究可望为脑-机交互控制/脑控机器人接口提供更多的力控制意图指令,奠定了后续的对比研究基础.     

正交拉普拉斯语种识别方法

提出了一种正交拉普拉斯语种识别方法,即在提取语音的i-vector后,采用正交局部保持投影进行子空间映射,将信号整体空间映射到语言信息加信道信息子空间,然后对映射后的矢量进行信道补偿处理,最后用支持向量机进行识别. 尽管i-vector最大限度地保留了语音的声学信息,但是并没有发现这些信息之间的内在结构. 利用正交局部保持投影在去除声学无关信息的基础上,进一步发现声学特征的内在结构,能够有效地提高特征的区分性. 在对NIST LRE 2003测试数据库实验后,发现新方法相较于基线系统来说,平均代价降低了28.91%.     

融合异构特征的子空间迁移学习算法

特征迁移重在领域共有特征间学习,然而其忽略领域特有特征的判别信息,使算法的适应性受到一定的局限. 针对此问题,提出了一种融合异构特征的子空间迁移学习（The subspace transfer learning algorithm integrating with heterogeneous features,STL-IHF）算法.该算法将数据的特征空间看成共享和特有两个特征子空间的组合,同时基于经验风险最 小框架将共享特征和特有特征共同嵌入到支持向量机（Support vector machine,SVM）的训练过程中.其在共享特征子空间上实现知识迁移的 同时兼顾了领域特有的异构信息,增强了算法的适应性.模拟和真实数据集上的实验结果表明了所提方法的有效性.     

BiELL: A bisection ELLPACK-based storage format for optimizing SpMV on GPUs

Sparse matrix–vector multiplication (SpMV) is one of the most important high level operations for basic linear algebra. Nowadays, the GPU has evolved into a highly parallel coprocessor which is suited to compute-intensive, highly parallel computation. Achieving high performance of SpMV on GPUs is relatively challenging, especially when the matrix has no specific structure. For these general sparse matrices, a new data structure based on the bisection ELLPACK format, BiELL, is designed to realize the load balance better, and thus improve the performance of the SpMV. Besides, based on the same idea of JAD format, the BiJAD format can be obtained. Experimental results on various matrices show that the BiELL and BiJAD formats perform better than other similar formats, especially when the number of non-zero elements per row varies a lot.     

Computer aided detection system for micro calcifications in digital mammograms

Breast cancer continues to be a significant public health problem in the world. Early detection is the key for improving breast cancer prognosis. Mammogram breast X-ray is considered the most reliable method in early detection of breast cancer. However, it is difficult for radiologists to provide both accurate and uniform evaluation for the enormous mammograms generated in widespread screening. Micro calcification clusters (MCCs) and masses are the two most important signs for the breast cancer, and their automated detection is very valuable for early breast cancer diagnosis. The main objective is to discuss the computer-aided detection system that has been proposed to assist the radiologists in detecting the specific abnormalities and improving the diagnostic accuracy in making the diagnostic decisions by applying techniques splits into three-steps procedure beginning with enhancement by using Histogram equalization (HE) and Morphological Enhancement, followed by segmentation based on Otsu's threshold the region of interest for the identification of micro calcifications and mass lesions, and at last classification stage, which classify between normal and micro calcifications ‘patterns and then classify between benign and malignant micro calcifications. In classification stage; three methods were used, the voting K-Nearest Neighbor classifier (K-NN) with prediction accuracy of 73%, Support Vector Machine classifier (SVM) with prediction accuracy of 83%, and Artificial Neural Network classifier (ANN) with prediction accuracy of 77%.     

Prediction of human breast and colon cancers from imbalanced data using nearest neighbor and support vector machines

This study proposes a novel prediction approach for human breast and colon cancers using different feature spaces. The proposed scheme consists of two stages: the preprocessor and the predictor. In the preprocessor stage, the mega-trend diffusion (MTD) technique is employed to increase the samples of the minority class, thereby balancing the dataset. In the predictor stage, machine-learning approaches of K-nearest neighbor (KNN) and support vector machines (SVM) are used to develop hybrid MTD-SVM and MTD-KNN prediction models. MTD-SVM model has provided the best values of accuracy, G-mean and Matthew's correlation coefficient of 96.71%, 96.70% and 71.98% for cancer/non-cancer dataset, breast/non-breast cancer dataset and colon/non-colon cancer dataset, respectively. We found that hybrid MTD-SVM is the best with respect to prediction performance and computational cost. MTD-KNN model has achieved moderately better prediction as compared to hybrid MTD-NB (Naïve Bayes) but at the expense of higher computing cost. MTD-KNN model is faster than MTD-RF (random forest) but its prediction is not better than MTD-RF. To the best of our knowledge, the reported results are the best results, so far, for these datasets. The proposed scheme indicates that the developed models can be used as a tool for the prediction of cancer. This scheme may be useful for study of any sequential information such as protein sequence or any nucleic acid sequence.