Selective negative correlation learning approach to incremental learning

Negative correlation learning (NCL) is a successful approach to constructing neural network ensembles. In batch learning mode, NCL outperforms many other ensemble learning approaches. Recently, NCL has also shown to be a potentially powerful approach to incremental learning, while the advantages of NCL have not yet been fully exploited. In this paper, we propose a selective NCL (SNCL) algorithm for incremental learning. Concretely, every time a new training data set is presented, the previously trained neural network ensemble is cloned. Then the cloned ensemble is trained on the new data set. After that, the new ensemble is combined with the previous ensemble and a selection process is applied to prune the whole ensemble to a fixed size. This paper is an extended version of our preliminary paper on SNCL. Compared to the previous work, this paper presents a deeper investigation into SNCL, considering different objective functions for the selection process and comparing SNCL to other NCL-based incremental learning algorithms on two more real world bioinformatics data sets. Experimental results demonstrate the advantage of SNCL. Further, comparisons between SNCL and other existing incremental learning algorithms, such Learn++ and ARTMAP, are also presented.     

基于标签正负相关性的多标签类属特征学习

提出一种基于标签正负相关性的多标签类属特征学习方法(multi-label learning with label-specific features based on positive and negative label correlation,LIFTPNL).基于k近邻的思想构建全局和局部的标签信息矩阵,根据此...     

Automatic identification of music performers with learning ensembles

This article addresses the problem of identifying the most likely music performer, given a set of performances of the same piece by a number of skilled candidate pianists. We propose a set of very simple features for representing stylistic characteristics of a music performer, introducing ‘norm-based’ features that relate to a kind of ‘average’ performance. A database of piano performances of 22 pianists playing two pieces by Frédéric Chopin is used in the presented experiments. Due to the limitations of the training set size and the characteristics of the input features we propose an ensemble of simple classifiers derived by both subsampling the training set and subsampling the input features. Experiments show that the proposed features are able to quantify the differences between music performers. The proposed ensemble can efficiently cope with multi-class music performer recognition under inter-piece conditions, a difficult musical task, displaying a level of accuracy unlikely to be matched by human listeners (under similar conditions).     

Combining features of negative correlation learning with mixture of experts in proposed ensemble methods

Both theoretical and experimental studies have shown that combining accurate neural networks (NNs) in the ensemble with negative error correlation greatly improves their generalization abilities. Negative correlation learning (NCL) and mixture of experts (ME), two popular combining methods, each employ different special error functions for the simultaneous training of NNs to produce negatively correlated NNs. In this paper, we review the properties of the NCL and ME methods, discussing their advantages and disadvantages. Characterization of both methods showed that they have different but complementary features, so if a hybrid system can be designed to include features of both NCL and ME, it may be better than each of its basis approaches. In this study, two approaches are proposed to combine the features of both methods in order to solve the weaknesses of one method with the strength of the other method, i.e., gated-NCL (G-NCL) and mixture of negatively correlated experts (MNCE). In the first approach, G-NCL, a dynamic combiner of ME is used to combine the outputs of base experts in the NCL method. The suggested combiner method provides an efficient tool to evaluate and combine the NCL experts by the weights estimated dynamically from the inputs based on the different competences of each expert regarding different parts of the problem. In the second approach, MNCE, the capability of a control parameter for NCL is incorporated in the error function of ME, which enables the training algorithm of ME to efficiently adjust the measure of negative correlation between the experts. This control parameter can be regarded as a regularization term added to the error function of ME to establish better balance in bias–variance–covariance trade-offs and thus improves the generalization ability. The two proposed hybrid ensemble methods, G-NCL and MNCE, are compared with their constituent methods, ME and NCL, in solving several benchmark problems. The experimental results show that our proposed methods preserve the advantages and alleviate the disadvantages of their basis approaches, offering significantly improved performance over the original methods.     

Evolutionary extreme learning machine

Extreme learning machine (ELM) [G.-B. Huang, Q.-Y. Zhu, C.-K. Siew, Extreme learning machine: a new learning scheme of feedforward neural networks, in: Proceedings of the International Joint Conference on Neural Networks (IJCNN2004), Budapest, Hungary, 25-29 July 2004], a novel learning algorithm much faster than the traditional gradient-based learning algorithms, was proposed recently for single-hidden-layer feedforward neural networks (SLFNs). However, ELM may need higher number of hidden neurons due to the random determination of the input weights and hidden biases. In this paper, a hybrid learning algorithm is proposed which uses the differential evolutionary algorithm to select the input weights and Moore-Penrose (MP) generalized inverse to analytically determine the output weights. Experimental results show that this approach is able to achieve good generalization performance with much more compact networks.     

Discovering predictive ensembles for transfer learning and meta-learning

Recent meta-learning approaches are oriented towards algorithm selection, optimization or recommendation of existing algorithms. In this article we show how data-tailored algorithms can be constructed from building blocks on small data sub-samples. Building blocks, typically weak learners, are optimized and evolved into data-tailored hierarchical ensembles. Good-performing algorithms discovered by evolutionary algorithm can be reused on data sets of comparable complexity. Furthermore, these algorithms can be scaled up to model large data sets. We demonstrate how one particular template (simple ensemble of fast sigmoidal regression models) outperforms state-of-the-art approaches on the Airline data set. Evolved hierarchical ensembles can therefore be beneficial as algorithmic building blocks in meta-learning, including meta-learning at scale.     

Evolutionary learning of nearest-neighbor MLP

The nearest-neighbor multilayer perceptron (NN-MLP) is a single-hidden-layer network suitable for pattern recognition. To design an NN-MLP efficiently, this paper proposes a new evolutionary algorithm consisting of four basic operations: recognition, remembrance, reduction, and review. Experimental results show that this algorithm can produce the smallest or nearly smallest networks from random initial ones.     

Dynamically building diversified classifier pruning ensembles via canonical correlation analysis

Empirical studies on ensemble learning that combines multiple classifiers have shown that, it is an effective technique to improve accuracy and stability of a single classifier. In this paper, we propose a novel method of dynamically building diversified sparse ensembles. We first apply a technique known as the canonical correlation to model the relationship between the input data variables and output base classifiers. The canonical (projected) output classifiers and input training data variables are encoded globally through a multi-linear projection of CCA, to decrease the impacts of noisy input data and incorrect classifiers to a minimum degree in such a global view. Secondly, based on the projection, a sparse regression method is used to prune representative classifiers by combining classifier diversity measurement. Based on the above methods, we evaluate the proposed approach by several datasets, such as UCI and handwritten digit recognition. Experimental results of the study show that, the proposed approach achieves better accuracy as compared to other ensemble methods such as QFWEC, Simple Vote Rule, Random Forest, Drep and Adaboost.

     

Evolutionary learning of hierarchical decision rules

This paper describes an approach based on evolutionary algorithms, hierarchical decision rules (HIDER), for learning rules in continuous and discrete domains. The algorithm produces a hierarchical set of rules, that is, the rules are sequentially obtained and must therefore be tried until one is found whose conditions are satisfied. Thus, the number of rules may be reduced because the rules could be inside of one another. The evolutionary algorithm uses both real and binary coding for the individuals of the population. We tested our system on real data from the UCI repository, and the results of a ten-fold cross-validation are compared to C4.5s, C4.5Rules, See5s, and See5Rules. The experiments show that HIDER works well in practice.     

Unsupervised learning of arbitrarily shaped clusters using ensembles of Gaussian models

We propose a new clustering algorithm, called SyMP, which is based on synchronization of pulse-coupled oscillators. SyMP represents each data point by an Integrate-and-Fire oscillator and uses the relative similarity between the points to model the interaction between the oscillators. SyMP is robust to noise and outliers, determines the number of clusters in an unsupervised manner, and identifies clusters of arbitrary shapes. The robustness of SyMP is an intrinsic property of the synchronization mechanism. To determine the optimum number of clusters, SyMP uses a dynamic and cluster dependent resolution parameter. To identify clusters of various shapes, SyMP models each cluster by an ensemble of Gaussian components. SyMP does not require the specification of the number of components for each cluster. This number is automatically determined using a dynamic intra-cluster resolution parameter. Clusters with simple shapes would be modeled by few components while clusters with more complex shapes would require a larger number of components. The proposed clustering approach is empirically evaluated with several synthetic data sets, and its performance is compared with GK and CURE. To illustrate the performance of SyMP on real and high-dimensional data sets, we use it to categorize two image databases.     

Feature elimination based random subspace ensembles learning for ECG arrhythmia diagnosis

Various methods for ensembles selection and classifier combination have been designed to optimize the performance of ensembles of classifiers. However, use of large number of features in training data can affect the classification performance of machine learning algorithms. The objective of this paper is to represent a novel feature elimination (FE) based ensembles learning method which is an extension to an existing machine learning environment. Here the standard 12 lead ECG signal recordings data have been used in order to diagnose arrhythmia by classifying it into normal and abnormal subjects. The advantage of the proposed approach is that it reduces the size of feature space by way of using various feature elimination methods. The decisions obtained from these methods have been coalesced to form a fused data. Thus the idea behind this work is to discover a reduced feature space so that a classifier built using this tiny data set would perform no worse than a classifier built from the original data set. Random subspace based ensembles classifier is used with PART tree as base classifier. The proposed approach has been implemented and evaluated on the UCI ECG signal data. Here, the classification performance has been evaluated using measures such as mean absolute error, root mean squared error, relative absolute error, F-measure, classification accuracy, receiver operating characteristics and area under curve. In this way, the proposed novel approach has provided an attractive performance in terms of overall classification accuracy of 91.11 % on unseen test data set. From this work, it is shown that this approach performs well on the ensembles size of 15 and 20.     

Negative correlation in incremental learning

Negative Correlation Learning (NCL) has been successfully applied to construct neural network ensembles. It encourages the neural networks that compose the ensemble to be different from each other and, at the same time, accurate. The difference among the neural networks that compose an ensemble is a desirable feature to perform incremental learning, for some of the neural networks can be able to adapt faster and better to new data than the others. So, NCL is a potentially powerful approach to incremental learning. With this in mind, this paper presents an analysis of NCL, aiming at determining its weak and strong points to incremental learning. The analysis shows that it is possible to use NCL to overcome catastrophic forgetting, an important problem related to incremental learning. However, when catastrophic forgetting is very low, no advantage of using more than one neural network of the ensemble to learn new data is taken and the test error is high. When all the neural networks are used to learn new data, some of them can indeed adapt better than the others, but a higher catastrophic forgetting is obtained. In this way, it is important to find a trade-off between overcoming catastrophic forgetting and using an entire ensemble to learn new data. The NCL results are comparable with other approaches which were specifically designed to incremental learning. Thus, the study presented in this work reveals encouraging results with negative correlation in incremental learning, showing that NCL is a promising approach to incremental learning.

Unsupervised feature selection using clustering ensembles and population based incremental learning algorithm

This paper describes a novel feature selection algorithm for unsupervised clustering, that combines the clustering ensembles method and the population based incremental learning algorithm. The main idea of the proposed unsupervised feature selection algorithm is to search for a subset of all features such that the clustering algorithm trained on this feature subset can achieve the most similar clustering solution to the one obtained by an ensemble learning algorithm. In particular, a clustering solution is firstly achieved by a clustering ensembles method, then the population based incremental learning algorithm is adopted to find the feature subset that best fits the obtained clustering solution. One advantage of the proposed unsupervised feature selection algorithm is that it is dimensionality-unbiased. In addition, the proposed unsupervised feature selection algorithm leverages the consensus across multiple clustering solutions. Experimental results on several real data sets demonstrate that the proposed unsupervised feature selection algorithm is often able to obtain a better feature subset when compared with other existing unsupervised feature selection algorithms.     

Evolutionary learning of rule premises for fuzzy modelling

The task of fuzzy modelling involves specification of rule antecedents and determination of their consequent counterparts. Rule premises appear here a critical issue since they determine the structure of a rule base. This paper proposes a new approach to extracting fuzzy rules from training examples by means of genetic-based premise learning. In order to construct a 'parsimonious' fuzzy model with high generalization ability, general premise structure allowing incomplete compositions of input variables as well as OR connectives of linguistic terms is considered. A genetic algorithm is utilized to optimize both the premise structure of rules and fuzzy set membership functions at the same time. Determination of rule conclusions is nested in the premise learning, where consequences of individual rules are determined under fixed preconditions. The proposed method was applied to the well-known gas furnace data of Box and Jenkins to show its validity and to compare its performance with those of other works.     

Evolutionary selection extreme learning machine optimization for regression

Neural network model of aggression can approximate unknown datasets with the less error. As an important method of global regression, extreme learning machine (ELM) represents a typical learning method in single-hidden layer feedforward network, because of the better generalization performance and the faster implementation. The “randomness” property of input weights makes the nonlinear combination reach arbitrary function approximation. In this paper, we attempt to seek the alternative mechanism to input connections. The idea is derived from the evolutionary algorithm. After predefining the number L of hidden nodes, we generate original ELM models. Each hidden node is seemed as a gene. To rank these hidden nodes, the larger weight nodes are reassigned for the updated ELM. We put L/2 trivial hidden nodes in a candidate reservoir. Then, we generate L/2 new hidden nodes to combine L hidden nodes from this candidate reservoir. Another ranking is used to choose these hidden nodes. The fitness-proportional selection may select L/2 hidden nodes and recombine evolutionary selection ELM. The entire algorithm can be applied for large-scale dataset regression. The verification shows that the regression performance is better than the traditional ELM and Bayesian ELM under less cost gain.     

Tune and mix: learning to rank using ensembles of calibrated multi-class classifiers

In subset ranking, the goal is to learn a ranking function that approximates a gold standard partial ordering of a set of objects (in our case, a set of documents retrieved for the same query). The partial ordering is given by relevance labels representing the relevance of documents with respect to the query on an absolute scale. Our approach consists of three simple steps. First, we train standard multi-class classifiers (AdaBoost.MH and multi-class SVM) to discriminate between the relevance labels. Second, the posteriors of multi-class classifiers are calibrated using probabilistic and regression losses in order to estimate the Bayes-scoring function which optimizes the Normalized Discounted Cumulative Gain (NDCG). In the third step, instead of selecting the best multi-class hyperparameters and the best calibration, we mix all the learned models in a simple ensemble scheme. Our extensive experimental study is itself a substantial contribution. We compare most of the existing learning-to-rank techniques on all of the available large-scale benchmark data sets using a standardized implementation of the NDCG score. We show that our approach is competitive with conceptually more complex listwise and pairwise methods, and clearly outperforms them as the data size grows. As a technical contribution, we clarify some of the confusing results related to the ambiguities of the evaluation tools, and propose guidelines for future studies.     

Asymmetry label correlation for multi-label learning

As an effective method for mining latent information between labels, label correlation is widely adopted by many scholars to model multi-label learning algorithms. Most existing multi-label algorithms usually ignore that the correlation between labels may be asymmetric while asymmetry correlation commonly exists in the real-world scenario. To tackle this problem, a multi-label learning algorithm with asymmetry label correlation (ACML, Asymmetry Label Correlation for Multi-Label Learning) is proposed in this paper. First, measure the adjacency between labels to construct the label adjacency matrix. Then, cosine similarity is utilized to construct the label correlation matrix. Finally, we constrain the label correlation matrix with the label adjacency matrix. Thus, asymmetry label correlation is modeled for multi-label learning. Experiments on multiple multi-label benchmark datasets show that the ACML algorithm has certain advantages over other comparison algorithms. The results of statistical hypothesis testing further illustrate the effectiveness of the proposed algorithm.

     

Canonical correlation analysis: an overview with application to learning methods

We present a general method using kernel canonical correlation analysis to learn a semantic representation to web images and their associated text. The semantic space provides a common representation and enables a comparison between the text and images. In the experiments, we look at two approaches of retrieving images based on only their content from a text query. We compare orthogonalization approaches against a standard cross-representation retrieval technique known as the generalized vector space model.     

基于误差相关度学习样本选择

针对有限样本学习机器的偏差/方差的困境,以及过拟合引起的泛化性能的下降,分析了样本选择对学习机器泛化的影响,提出误差相关度学习算法ECL,利用误差相关度来权衡偏差和方差的关系,避免了求解复杂学习系统的VC维数,并以样本点的误差相关度为指标来选择训练子集,提高学习机器的泛化性能.仿真结果表明ECL算法有效地抑制过拟合现象的发生,保证学习机器泛化性能的提高.     

Evolutionary learning with a neuromolecular architecture: a biologically motivated approach to computational adaptability

 The effectiveness of evolutionary learning depends both on the variation-selection search operations used and on the structure-function relations of the organization to which these operations are applied. Some organizations—in particular those that occur in biology—are more evolution friendly than others. We describe an artificial neuromolecular (ANM) architecture that illustrates the structure-function relationships that underlie evolutionary adaptability and the manner in which these relationships can be represented in computer programs. The ANM system, a brain-like design that combines intra- and interneuronal levels of processing, can be coupled to a variety of pattern recognition-effector control tasks. The capabilities of the model, in particular its adaptability properties, are here illustrated in the context of Chinese character recognition.