MKL-SING: A data-driven approach of sign recognition for managing and improving public services

The detection and identification of traffic signs is a fundamental function of an intelligent transportation system. The extraction or identification of a road sign poses the same problems as object identification in natural contexts: conditions of illumination are variable and uncontrollable, and various objects frequently surround road signs. These difficulties make the extraction of features difficult. The fusion of time and space features of traffic signs is important for improving the performance of sign recognition. Deep learning-based algorithms are time-consuming to train based on a large amount of data. They are difficult to deploy on resource-constrained portable devices and conduct sign detection in real time. The accuracy of sign detection should be further improved, which is related to the safety of traffic participants. To improve the accuracy of feature extraction and classification of traffic signs, we propose MKL-SING, a hybrid approach based on multi-kernel support vector machine (MKL-SVM) for public transportation SIGN recognition. It contains three main components: a principal component analysis for image dimension reduction, a fused feature extractor, and a multi-kernel SVM-based classifier. The fused feature extractor extracts and fuses the time and space features of traffic signs. The multi-kernel SVM then classifies the traffic signs based on the fused features. Different kernel functions in the multi-kernel SVM are fused based on a feature weighting procedure. Compared with single-core SVM, multi-kernel SVM can better process massive data because it can project each kernel function into high-dimensional feature space to get global solutions. Finally, the performance of SVM-TSR is validated based on three traffic sign datasets. Experiment results show that SVM-TSR performs better than state-of-the-art methods in terms of dynamic traffic sign identification and recognition.     

Advanced learning algorithms for cross-language patent retrieval and classification

We study several machine learning algorithms for cross-language patent retrieval and classification. In comparison with most of other studies involving machine learning for cross-language information retrieval, which basically used learning techniques for monolingual sub-tasks, our learning algorithms exploit the bilingual training documents and learn a semantic representation from them. We study Japanese–English cross-language patent retrieval using Kernel Canonical Correlation Analysis (KCCA), a method of correlating linear relationships between two variables in kernel defined feature spaces. The results are quite encouraging and are significantly better than those obtained by other state of the art methods. We also investigate learning algorithms for cross-language document classification. The learning algorithm are based on KCCA and Support Vector Machines (SVM). In particular, we study two ways of combining the KCCA and SVM and found that one particular combination called SVM_2k achieved better results than other learning algorithms for either bilingual or monolingual test documents.     

Noise reduction through summarization for Web-page classification

Due to a large variety of noisy information embedded in Web pages, Web-page classification is much more difficult than pure-text classification. In this paper, we propose to improve the Web-page classification performance by removing the noise through summarization techniques. We first give empirical evidence that ideal Web-page summaries generated by human editors can indeed improve the performance of Web-page classification algorithms. We then put forward a new Web-page summarization algorithm based on Web-page layout and evaluate it along with several other state-of-the-art text summarization algorithms on the LookSmart Web directory. Experimental results show that the classification algorithms (NB or SVM) augmented by any summarization approach can achieve an improvement by more than 5.0% as compared to pure-text-based classification algorithms. We further introduce an ensemble method to combine the different summarization algorithms. The ensemble summarization method achieves more than 12.0% improvement over pure-text based methods.     

An efficient smart data mining framework based cloud internet of things for developing artificial intelligence of marketing information analysis

Because of the big volume of marketing data, a human analyst would be unable to uncover any useful information for marketing that could aid in the process of making decision. Smart Data Mining (SDM), which is considered an important field from Artificial Intelligence (AI) is completely assisting in the performance business management analytics and marketing information. In this study, most reliable six algorithms in SDM are applied; Naïve Bayes (NB), Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM), ID3, and C4.5 on actual data of marketing for bank that taken from Cloud Internet of Thing (CIoT). The objectives of this study are to build an efficient framework to increase campaign of marketing for banks by identifying main characteristics that affect a success and to test the performance of CIoT and SDM algorithms. This study is expected to enhance the scientific contributions to investigating the marketing information capacities by integrating SDM with CIoT. The performances of SDM algorithms are calculated by eight measures; accuracy, balance accuracy, precision, mean absolute error, root mean absolute error, recall, F1- Score and running time. The experimental findings show that the proposed framework is successful, with higher accuracies and good performance. Results revealed that customer service & marketing tactics are essential for a Company’ success & survival. Also, the C4.5 has accomplished better achievement than the SVM, RF, LR, NB, & ID3. At the end, CIoT Platform was evaluated by response time, request rate & processing of bank data.     

Support vector machines: relevance feedback and information retrieval

We compare support vector machines (SVMs) to Rocchio, Ide regular and Ide dec-hi algorithms in information retrieval (IR) of text documents using relevancy feedback. It is assumed a preliminary search finds a set of documents that the user marks as relevant or not and then feedback iterations commence. Particular attention is paid to IR searches where the number of relevant documents in the database is low and the preliminary set of documents used to start the search has few relevant documents. Experiments show that if inverse document frequency (IDF) weighting is not used because one is unwilling to pay the time penalty needed to obtain these features, then SVMs are better whether using term-frequency (TF) or binary weighting. SVM performance is marginally better than Ide dec-hi if TF-IDF weighting is used and there is a reasonable number of relevant documents found in the preliminary search. If the preliminary search is so poor that one has to search through many documents to find at least one relevant document, then SVM is preferred.     

A3CRank: An adaptive ranking method based on connectivity,content and click-through data

Due to the proliferation and abundance of information on the web, ranking algorithms play an important role in web search. Currently, there are some ranking algorithms based on content and connectivity such as BM25 and PageRank. Unfortunately, these algorithms have low precision and are not always satisfying for users. In this paper, we propose an adaptive method, called A3CRank, based on the content, connectivity, and click-through data triple. Our method tries to aggregate ranking algorithms such as BM25, PageRank, and TF-IDF. We have used reinforcement learning to incorporate user behavior and find a measure of user satisfaction for each ranking algorithm. Furthermore, OWA, an aggregation operator is used for merging the results of the various ranking algorithms. A3CRank adapts itself with user needs and makes use of user clicks to aggregate the results of ranking algorithms. A3CRank is designed to overcome some of the shortcomings of existing ranking algorithms by combining them together and producing an overall better ranking criterion. Experimental results indicate that A3CRank outperforms other combinational ranking algorithms such as Ranking SVM in terms of P@n and NDCG metrics. We have used 130 queries on University of California at Berkeley’s web to train and evaluate our method.     

支持向量机在分类问题中的应用研究

支持向量机是一种基于统计学习理论的机器学习方法,针对小样本情况表现出了优良的性能,目前被广泛应用于模式识别、函数回归、故障诊断等方面。这里主要研究支持向量机分类问题,着重讨论了以下几个方面的内容。首先介绍了支持向量机分类器算法,并将其应用于数据分类,取得了较高的准确率,所用数据来自于UCI数据集。仿真结果表明该算法具有较快的收敛速度和较高的计算精度。     

基于KNN和SVM的中文文本自动分类研究

借助文本分类系统软件,采用来自10个大类的中文文本数据,按照训练集与测试集2:1的比例,使用KNN和SVM分类算法,对数据集进行自动分类的实验。旨在通过具体的语料库实验,探讨文本自动分类的关键技术,分析、比较与评价实验结果,探讨文本分类中具体参数的设置和不同分类算法之优劣。     

Differential diagnosis of hepatocellular carcinoma and hepatic hemangioma based on maximum wavelet-coefficient statistics: Novel radiomics features from plain CT

In computed tomography (CT)-based diagnoses of liver tumors, contrast-enhanced CT may cause renal toxicity and allergic reactions. Regular health examinations prefer plain CT, but subsequent diagnoses significantly depend on subjective experience. Radiomics provides a quantitative, objective, and noninvasive way for diagnosing liver tumors. This study aimed to use plain CT-based radiomics to diagnose hepatocellular (HCC, malignant) and hemangioma (HH, benign) liver tumors. Inspired by the knowledge that HCC and HH exhibit different histopathological characteristics, we developed a novel feature extraction technique (referred to as maximum wavelet-coefficient statistics, MWCS) to highlight the differences in histopathological characteristics by reorganizing and expressing the patterns of wavelet-coefficients that represent local changes. We attempted multiple feature selection algorithms and various machine learning approaches to train classification models and tested these models on an independent test cohort. Experimental results showed that the classification models based on the proposed MWCS-COM (using a statistical method of co-occurrence matrix in MWCS) feature set exhibited performance superior to those based on traditional feature sets. Furthermore, the linear support vector machine (SVM) model achieved state-of-the-art performance in the classification experiments with a test area under receiver operator characteristic curve (AUC) of 0.8734 (95% confidence interval, 0.8666–0.8802). This result indicated that the MWCS-COM features are highly advantageous to the differential diagnosis of HCC and HH from plain CT images. We also explored the potential associations between MWCS-COM features and histopathological characteristics and observed that the MWCS-COM features could potentially enhance radiologists’ diagnostic ability.     

Fast implementation of two-dimensional singular spectrum analysis for effective data classification in hyperspectral imaging

Although singular spectrum analysis (SSA) has been successfully applied for data classification in hyperspectral remote sensing, it suffers from extremely high computational cost, especially for 2D-SSA. As a result, a fast implementation of 2D-SSA namely F-2D-SSA is presented in this paper, where the computational complexity has been significantly reduced with a rate up to 60%. From comprehensive experiments undertaken, the effectiveness of F-2D-SSA is validated producing a similar high-level of accuracy in pixel classification using support vector machine (SVM) classifier, yet with a much reduced complexity in comparison to conventional 2D-SSA. Therefore, the introduction and evaluation of F-2D-SSA completes a series of studies focused on SSA, where in this particular research, the reduction in computational complexity leads to potential applications in mobile and embedded devices such as airborne or satellite platforms.     

An intelligent propagation distance estimation algorithm based on fundamental frequency energy distribution for periodic vibration localization

Earth surface vibrations generated by passing vehicles, excavation equipment, footsteps, etc., attract increasing attentions in the research community due to their wide applications. In this paper, we investigate the periodic vibration source localization problem, which has recently shown significance in excavation device detection and localization for urban underground pipeline network protection. An intelligent propagation distance estimation algorithm based on a novel fundamental frequency energy distribution (FBED) feature is developed for periodic vibration signal localization. Contributions of the paper lie in three aspects: 1) a novel frequency band energy distribution (FBED) feature is developed to characterize the property of vibrations at different propagation distances; 2) an intelligent propagation distance estimation model built on the FBED feature with machine learning algorithms is proposed, where for comparisons, the support vector machine (SVM) for regression and regularized extreme learning machine (RELM) are used; 3) a localization algorithm based on the distance-of-arrival (DisOA) estimation using three piezoelectric transducer sensors is given for source position estimation. To testify the effectiveness of the proposed algorithms, case studies on real collected periodic vibration signals generated by two electric hammers with different fundamental frequencies are presented in the paper. The transmission medium is the cement road and experiments on vibration signals recorded at different propagation distances are conducted.     

网络舆情安全应用中主题分类方法的研究与实现

针对网络舆情安全应用中主题分类问题展开研究。主题分类需要解决5个问题。建立了网络舆情安全应用的分类体系;介绍了文档特征提取和选择方法;分析了向量空间模型;实现了KNN和SVM两种分类算法。验证了方法的实用性。     

Artificial intelligence implication on energy sustainability in Internet of Things: A survey

The massive number of Internet of Things (IoT) devices connected to the Internet is continuously increasing. The operations of these devices rely on consuming huge amounts of energy. Power limitation is a major issue hindering the operation of IoT applications and services. To improve operational visibility, Low-power devices which constitute IoT networks, drive the need for sustainable sources of energy to carry out their tasks for a prolonged period of time. Moreover, the means to ensure energy sustainability and QoS must consider the stochastic nature of the energy supplies and dynamic IoT environments. Artificial Intelligence (AI) enhanced protocols and algorithms are capable of predicting and forecasting demand as well as providing leverage at different stages of energy use to supply. AI will improve the efficiency of energy infrastructure and decrease waste in distributed energy systems, ensuring their long-term viability. In this paper, we conduct a survey to explore enhanced AI-based solutions to achieve energy sustainability in IoT applications. AI is relevant through the integration of various Machine Learning (ML) and Swarm Intelligence (SI) techniques in the design of existing protocols. ML mechanisms used in the literature include variously supervised and unsupervised learning methods as well as reinforcement learning (RL) solutions. The survey constitutes a complete guideline for readers who wish to get acquainted with recent development and research advances in AI-based energy sustainability in IoT Networks. The survey also explores the different open issues and challenges.     

关于支持向量回归机的模型选择

支持向量机是在统计学习理论基础上发展起来的一种新型的机器学习方法。模型选择是设计支持向量机的重要内容之一。本文在分析用于回归的支持向量机原理的基础上,分别从核函数的选择、模型参数的作用、模型参数的调整方法等模型选择方面进行了综述,并讨论了模型选择的优缺点,最后指出在实际应用中常见的核函数和模型参数调整方法。     

Early prediction of cognitive impairments using physiological signal for enhanced socioeconomic status

Cognitive impairments like memory disorder and depressive disorders lead to fatal consequences if proper attention is not given to such health hazards. Their impact is extended to the socioeconomic status of the developed and low or middle-income countries in terms of loss of talented and skilled population. Additionally, financial burden is borne by the countries in terms of additional health budget allotment. This paper presents a novel strategy for early detection of cognitive deficiency to eliminate the economic repercussions caused by memory disorder and depressive disorders. In this work, Electroencephalogram (EEG) and a word learning neuropsychological test, i.e. California Verbal Learning Task (CVLT), are conjunctively used for memory assessment. The features of EEG and scores of CVLT are modeled by applying different machine learning techniques, namely K-Nearest Neighbor (KNN), Gaussian Naive Bayes (GNB), Decision Tree (DT), Random Forest (RF), and Support Vector Machine (SVM). Comparatively, experimental results have better classification accuracy than the existing schemes that considered EEG for estimating cognitive heuristics. More specifically, SVM attains the highest accuracy score of 81.56% among all machine learning algorithms, which can assist in the early detection of cognitive impairments. The proposed strategy can be helpful in clinical diagnosis of psychological health and improving quality of life as a whole.     

SVM决策树在沉积微相识别中的应用

支持向量机(SVM)作为统计学理论最年轻的分支,其应用日益广泛。针对油层沉积微相的多类识别问题,可采用支持向量机和决策树相结合的方法。对传统的SVM决策树进行改进的基础上,在SVM核函数选取过程中,构造了与实际问题有关的核函数。此方法有效的降低了支持向量机的设计难度,同时提高了识别精度和泛化能力。最后用实例对比神经网络验证了该方法的优越性。     

Few-shot entity linking of food names

Entity linking (EL), the task of automatically matching mentions in text to concepts in a target knowledge base, remains under-explored when it comes to the food domain, despite its many potential applications, e.g., finding the nutritional value of ingredients in databases. In this paper, we describe the creation of new resources supporting the development of EL methods applied to the food domain: the E.Care Knowledge Base (E.Care KB) which contains 664 food concepts and the E.Care dataset, a corpus of 468 cooking recipes where ingredient names have been manually linked to corresponding concepts in the E.Care KB. We developed and evaluated different methods for EL, namely, deep learning-based approaches underpinned by Siamese networks trained under a few-shot learning setting, traditional machine learning-based approaches underpinned by support vector machines (SVMs) and unsupervised approaches based on string matching algorithms. Combining the strengths of each of these approaches, we built a hybrid model for food EL that balances the trade-offs between performance and inference speed. Specifically, our hybrid model obtains 89.40% accuracy and links mentions at an average speed of 0.24 seconds per mention, whereas our best deep learning-based model, SVM model and unsupervised model obtain accuracies of 86.99%, 87.19% and 87.43% at inference speeds of 0.007, 0.66 and 0.02 seconds per mention, respectively.     

基于语义扩展的数字文献自动分类方法研究

针对图书、期刊论文等数字文献文本特征较少而导致特征向量语义表达不够准确、分类效果差的问题,本文提出一种基于特征语义扩展的数字文献分类方法。该方法首先利用TF-IDF方法获取对数字文献文本表示能力较强、具有较高TF-IDF值的核心特征词;其次分别借助知网(Hownet)语义词典以及开放知识库维基百科(Wikipedia)对核心特征词集进行语义概念的扩展,以构建维度较低、语义丰富的概念向量空间;最后采用MaxEnt、SVM等多种算法构造分类器实现对数字文献的自动分类。实验结果表明:相比传统基于特征选择的短文本分类方法,该方法能有效地实现对短文本特征的语义扩展,提高数字文献分类的分类性能。     

基于支持向量机优化参数的集成多核典型相关分析

提出一种可有效处理多个数据集合之间变量关系的集成多核典型相关分析方法,构造一个特殊的核函数,使其更好地将原始样本数据映射到高维空间;基于支持向量机,在选择一个优化参数的基础上最大化多组数据集变量间的关系,以寻求整体相关性最大。在多特征手写体数字库上的实验证明,相比传统的典型相关分析与核典型相关分析方法,基于优化参数的集成多核典型相关分析方法具有更优的性能。     

Developing scalable management information system with big financial data using data mart and mining architecture

The traditional Management Information System (MIS) with Big Financial Data (BFD) for corporate financial diagnosis has many limitations such as the data is not summarized thus these causing increases in query times, and also the complexity in analysis. The creation of a Data Mart (DM) leads to a great summarization of data, such that contains only essential business information. And by using data mining techniques we can be extracting unknown useful information from DM and apply it to make important decisions for the business. Thus, in this paper we are adopting an architecture of six layers; interface layer, analysis layer, extract transformation load layer, data mart layer, data mining layer, and evaluating layer, MIS with BFD using DM and Mining (MIS-BFD-DMM) is proposed, which is not only permits the use of DM and mining technologies in decision support, but also the full utilization of non-financial/financial info held by businesses. This paper offers the benefits of building and integrating DM with mining. Also determines the distinction between DM and a relational database for decision-makers to get information. The test and analysis are achieved in the terms of useful metrics (accuracy, balance accuracy, F-measure, precision, recall, and time). As a result, Data returned from arranged star schema is far faster than ERD. In conclusion, the SVM is best than other algorithms in terms of the parameters of the confusion matrix.