主动协同半监督粗糙集分类模型

粗糙集理论是一种有监督学习模型,一般需要适量有标记的数据来训练分类器。但现实一些问题往往存在大量无标记的数据,而有标记数据由于标记代价过大较为稀少。文中结合主动学习和协同训练理论,提出一种可有效利用无标记数据提升分类性能的半监督粗糙集模型。该模型利用半监督属性约简算法提取两个差异性较大的约简构造基分类器,然后基于主动学习思想在无标记数据中选择两分类器分歧较大的样本进行人工标注,并将更新后的分类器交互协同学习。UCI数据集实验对比分析表明,该模型能明显提高分类学习性能,甚至能达到数据集的最优值。     

不完备弱标记数据的粗糙协同学习模型

针对不完备弱标记数据的学习问题,提出基于粗糙集理论的半监督协同学习模型.首先定义不完备弱标记数据的半监督差别矩阵,提出充分、具有差异性的约简子空间获取算法.然后在有标记数据集上利用各约简子空间训练两个基分类器.在无标记数据上,各分类器基于协同学习的思想标注信度较大的无标记样本给另一分类器学习,迭代更新直至无可利用的无标记数据.UCI数据集实验对比分析表明,文中模型可以获得更好的不完备弱标记数据的分类学习性能,具有有效性.     

基于单类分类器的半监督学习

提出一种结合单类学习器和集成学习优点的Ensemble one-class半监督学习算法.该算法首先为少量有标识数据中的两类数据分别建立两个单类分类器.然后用建立好的两个单类分类器共同对无标识样本进行识别,利用已识别的无标识样本对已建立的两个分类面进行调整、优化.最终被识别出来的无标识数据和有标识数据集合在一起训练一个基分类器,多个基分类器集成在一起对测试样本的测试结果进行投票.在5个UCI数据集上进行实验表明,该算法与tri-training算法相比平均识别精度提高4.5%,与仅采用纯有标识数据的单类分类器相比,平均识别精度提高8.9%.从实验结果可以看出,该算法在解决半监督问题上是有效的.     

基于辅助学习与富信息策略的Tri-training算法

针对Tri-training算法利用无标记样例时会引入噪声且限制无标记样例的利用率而导致分类性能下降的缺点,提出了AR-Tri-training(Tri-training with assistant and rich strategy)算法.提出辅助学习策略,结合富信息策略设计辅助学习器,并将辅助学习器应用在Tri-training训练以及说话声识别中.实验结果表明,辅助学习器在Tri-training训练的基础上不仅降低每次迭代可能产生的误标记样例数,而且能够充分地利用无标记样例以及在验证集上的错分样例信息.从实验结果可以得出,该算法能够弥补Tri-training算法的缺点,进一步提高测试率.     

基于多视图Tri-Training的微博用户性别判断

互联网技术不断发展,新浪微博作为公开的网络社交平台拥有庞大的活跃用户. 然而由于用户数量庞大,且个人信息并不一定真实,造成训练样本打标困难. 本文采用了一种多视图tri-training的方法,构建三个不同的视图,利用这些视图中少量已打标样本和未打标样本不断重复互相训练三个不同的分类器,最后集成这三个分类器实现用户性别判断. 本文用真实用户数据进行实验,发现和单一视图分类器相比,使用多视图tri-training学习训练后的分类器准确性更好,且需要打标的样本更少.     

一种基于树型贝叶斯网络的集成多标记分类算法

在多标记分类问题中,有效地利用标记间的依赖关系是进一步提升分类器性能的主要途径之一。基于分类器链算法,利用互信息度量理论构造分类对象的类属性之间明确的多标记关系依赖模型,并依据建立的标记依赖模型将分类器链中的线性依赖拓展成树型依赖,以适应更为复杂的标记依赖关系;同时,在此基础上利用Stacking集成学习方法建立最终训练模型,提出了一种新的针对树型依赖表示模型的Stacking算法。 在多个实验数据集上的实验结果表明,与原有的Stacking集成学习相比,该算法提升了分类器的相应评价指标。     

基于分类不确定性最小化的半监督集成学习算法

半监督集成是将半监督学习与集成学习相结合的一种学习范式，它一方面通过无标记样本来提高集成学习的多样性，同时解决集成学习样本量不足的问题，另一方面集成多个分类器能够进一步提升半监督学习模型的性能。现有的研究从理论和实践两个角度证明了半监督学习与集成学习之间的互益性。针对当前半监督集成学习算法对无标记样本信息利用不完全的缺陷，文中提出了一种新的基于分类不确定性最小化的半监督集成学习(Classification Uncertainty Minimization-Based Semi-Supervised Ensemble Learning, CUM-SSEL)算法，它引入信息熵作为对无标记样本进行打标的置信度评判标准，通过最小化无标记样本打标过程中的不确定性迭代地训练分类器，实现对无标记样本的高效利用，以增强分类器的泛化性能。在标准的实验数据集上对CUM-SSEL算法的可行性、合理性和有效性进行了验证，实验表明：随着基分类器的增加，CUM-SSEL算法的训练呈现收敛的趋势，同时它能够获得优于Self-Training, Co-Training, Tri-Training, Semi-Boo...     

协同半监督的构造性学习方法

利用构造性学习（CML）算法训练分类器需要大量已标记样本,然而获取大量已标记的样本较为困难。为此,提出了一种协同半监督的构造性学习算法。将已标记样本等分为三个训练集,分别使用构造性学习算法训练三个单分类器,以共同投票的方式对未标记样本进行标记,从而依次扩充三个单分类器训练集直到不能再扩充为止。将三个训练集合并训练出最终的分类器。选取UCI数据集进行实验,结果表明,与CML算法、Tri-CML算法、NB算法及Tri-NB相比,该方法的分类更为有效。     

基于Tri-training的半监督多标记学习算法

传统的多标记学习是监督意义下的学习,它要求获得完整的类别标记.但是当数据规模较大且类别数目较多时,获得完整类别标记的训练样本集是非常困难的.因而,在半监督协同训练思想的框架下,提出了基于Tri-training的半监督多标记学习算法（SMLT）.在学习阶段,SMLT引入一个虚拟类标记,然后针对每一对类别标记,利用协同训练机制Tri-training算法训练得到对应的分类器;在预测阶段,给定一个新的样本,将其代入上述所得的分类器中,根据类别标记得票数的多少将多标记学习问题转化为标记排序问题,并将虚拟类标记的得票数作为阈值对标记排序结果进行划分.在UCI中4个常用的多标记数据集上的对比实验表明,SMLT算法在4个评价指标上的性能大多优于其他对比算法,验证了该算法的有效性.     

一种新型多标记懒惰学习算法

在多标记学习框架下,每个样本由单个实例进行表示并同时对应于多个概念标记.已有的多标记懒惰学习算法并未充分考察样本多个标记之间的相关性,因此其泛化性能将会受到一定程度的不利影响.针对上述问题,提出一种新型多标记懒惰学习算法IMLLA.该算法首先找出测试样本在训练集中与各个概念类对应的近邻样本,然后基于近邻样本的多标记信息构造一个标记计数向量,并提交给已训练的线性分类器进行预测.由于IMLLA在对每个概念类进行预测时利用了蕴含于其他概念类中的信息,因而充分考察了样本多个标记之间的相关性.在人工数据集以及真实世界数据集上的实验表明,IMLLA算法的性能显著优于常用的多标记学习算法.     

Scaling Up Inductive Logic Programming: An Evolutionary Wrapper Approach

Inductive logic programming (ILP) algorithms are classification algorithms that construct classifiers represented as logic programs. ILP algorithms have a number of attractive features, notably the ability to make use of declarative background (user-supplied) knowledge. However, ILP algorithms deal poorly with large data sets (>10⁴ examples) and their widespread use of the greedy set-covering algorithm renders them susceptible to local maxima in the space of logic programs.This paper presents a novel approach to address these problems based on combining the local search properties of an inductive logic programming algorithm with the global search properties of an evolutionary algorithm. The proposed algorithm may be viewed as an evolutionary wrapper around a population of ILP algorithms.The evolutionary wrapper approach is evaluated on two domains. The chess-endgame (KRK) problem is an artificial domain that is a widely used benchmark in inductive logic programming, and Part-of-Speech Tagging is a real-world problem from the field of Natural Language Processing. In the latter domain, data originates from excerpts of the Wall Street Journal. Results indicate that significant improvements in predictive accuracy can be achieved over a conventional ILP approach when data is plentiful and noisy.     

基于空间邻域信息的高光谱遥感影像半监督协同训练

针对tri_training协同训练算法在小样本的高光谱遥感影像半监督分类过程中,存在增选样本的误标记问题,提出一种基于空间邻域信息的半监督协同训练分类算法tri_training_SNI(tri_training based on Spatial Neighborhood Information)。首先利用分类器度量方法不一致度量和新提出的不一致精度度量从MLR(Multinomial Logistic Regression)、KNN(k-Nearest Neighbor)、ELM(Extreme Learning Machine)和RF(Random Forest)4个分类器中选择3分类性能差异性最大的3个分类器;然后在样本选择过程中,采用选择出来的3个分类器,在两个分类器分类结果相同的基础上,加入初始训练样本的8邻域信息进行未标记样本的二次筛选和标签的确定,提高了半监督学习的样本选择精度。通过对AVIRIS和ROSIS两景高光谱遥感影像进行分类实验,结果表明与传统的tri_training协同算法相比,该算法在分类精度方面有明显提高。     

基于集成学习的Self-training在入侵检测中的应用

针对入侵检测的标记数据难以获得的问题,提出一种基于集成学习的Self-training方法——正则化Self-training。该方法结合主动学习和正则化理论,利用无标记数据对已有的分类器（该分类器对分类模式已学习得很好）作进一步的改进。对三种主要的集成学习方法在不同标记数据比例下进行对比实验,实验结果表明：借助大量无标记数据可以改善组合分类器的分类边界,算法能显著地降低结果分类器的错误率。     

选择策略对遗传归纳逻辑程序设计收敛性能的影响

采用遗传算法（GA）作为归纳逻辑程序设计（ILP）的搜索策略，可以提高ILP方法的鲁棒性和适应性，文章简要叙述了对作者提出的遗传归纳逻辑程序设计（GILP）算法作的改进，测试了选择策略对GILP算法收敛性能的影响，采用不同的选择策略不会影响算法的最终收敛结果，但会产生不同的选择压力，导致算法具有不同的收敛速率。     

Introducing possibilistic logic in ILP for dealing with exceptions

In this paper we propose a new formalization of the inductive logic programming (ILP) problem for a better handling of exceptions. It is now encoded in first-order possibilistic logic. This allows us to handle exceptions by means of prioritized rules, thus taking lessons from non-monotonic reasoning. Indeed, in classical first-order logic, the exceptions of the rules that constitute a hypothesis accumulate and classifying an example in two different classes, even if one is the right one, is not correct. The possibilistic formalization provides a sound encoding of non-monotonic reasoning that copes with rules with exceptions and prevents an example to be classified in more than one class. The benefits of our approach with respect to the use of first-order decision lists are pointed out. The possibilistic logic view of ILP problem leads to an optimization problem at the algorithmic level. An algorithm based on simulated annealing that in one turn computes the set of rules together with their priority levels is proposed. The reported experiments show that the algorithm is competitive to standard ILP approaches on benchmark examples.     

A second order cone programming approach for semi-supervised learning

Semi-supervised learning (SSL) involves the training of a decision rule from both labeled and unlabeled data. In this paper, we propose a novel SSL algorithm based on the multiple clusters per class assumption. The proposed algorithm consists of two stages. In the first stage, we aim to capture the local cluster structure of the training data by using the k-nearest-neighbor (kNN) algorithm to split the data into a number of disjoint subsets. In the second stage, a maximal margin classifier based on the second order cone programming (SOCP) is introduced to learn an inductive decision function from the obtained subsets globally. For linear classification problems, once the kNN algorithm has been performed, the proposed algorithm trains a classifier using only the first and second order moments of the subsets without considering individual data points. Since the number of subsets is usually much smaller than the number of training points, the proposed algorithm is efficient for handling big data sets with a large amount of unlabeled data. Despite its simplicity, the classification performance of the proposed algorithm is guaranteed by the maximal margin classifier. We demonstrate the efficiency and effectiveness of the proposed algorithm on both synthetic and real-world data sets.     

Text Classification from Labeled and Unlabeled Documents using EM

This paper shows that the accuracy of learned text classifiers can be improved by augmenting a small number of labeled training documents with a large pool of unlabeled documents. This is important because in many text classification problems obtaining training labels is expensive, while large quantities of unlabeled documents are readily available.We introduce an algorithm for learning from labeled and unlabeled documents based on the combination of Expectation-Maximization (EM) and a naive Bayes classifier. The algorithm first trains a classifier using the available labeled documents, and probabilistically labels the unlabeled documents. It then trains a new classifier using the labels for all the documents, and iterates to convergence. This basic EM procedure works well when the data conform to the generative assumptions of the model. However these assumptions are often violated in practice, and poor performance can result. We present two extensions to the algorithm that improve classification accuracy under these conditions: (1) a weighting factor to modulate the contribution of the unlabeled data, and (2) the use of multiple mixture components per class. Experimental results, obtained using text from three different real-world tasks, show that the use of unlabeled data reduces classification error by up to 30%.     

Safe semi-supervised learning: a brief introduction

Semi-supervised learning constructs the predictive model by learning from a few labeled training examples and a large pool of unlabeled ones. It has a wide range of application scenarios and has attracted much attention in the past decades. However, it is noteworthy that although the learning performance is expected to be improved by exploiting unlabeled data, some empirical studies show that there are situations where the use of unlabeled data may degenerate the performance. Thus, it is advisable to be able to exploit unlabeled data safely. This article reviews some research progress of safe semi-supervised learning, focusing on three types of safeness issue: data quality, where the training data is risky or of low-quality;model uncertainty, where the learning algorithm fails to handle the uncertainty during training; measure diversity, where the safe performance could be adapted to diverse measures.     

基于多分类器协同学习的卷积神经网络训练算法

卷积神经网络(CNN)是一类重要的深度神经网络,然而其训练过程需要大量的已标记样本,从而限制了其实际应用。针对这一问题,分析了CNN分类器的协同学习过程,给出了基于迭代进化的分类器协同训练算法CAMC。该算法结合了CNN和多分类器协同训练的优势,首先采用不同的卷积核提取出多种样本特征以产生不同的CNN分类器;然后利用少量的已标记样本和大量的未标记样本对多个分类器进行协同训练,以持续提高分类性能。在人脸表情标准数据集上的实验结果表明,相对于传统的表情特征识别法LBP和Gabor,CAMC能够在分类过程中利用未标记样本持续实现性能提升,从而具有更高的分类准确率。     

Parallel and Sequential Algorithms for Data Mining Using Inductive Logic

Inductive logic is a research area in the intersection of machine learning and logic programming, and has been increasingly applied to data mining. Inductive logic studies learning from examples, within the framework provided by clausal logic. It provides a uniform and expressive means of representation: examples, background knowledge, and induced theories are all expressed in first-order logic. Such an expressive representation is computationally expensive, so it is natural to consider improving the performance of inductive logic data mining using parallelism. We present a parallelization technique for inductive logic, and implement a parallel version of a core inductive logic programming system: Progol. The technique provides perfect partitioning of computation and data access and communication requirements are small, so almost linear speedup is readily achieved. However, we also show why the information flow of the technique permits superlinear speedup over the standard sequential algorithm. Performance results on several datasets and platforms are reported. The results have wider implications for the design on parallel and sequential data-mining algorithms. Received 30 August 2000 / Revised 30 January 2001 / Accepted in revised form 16 May 2001