Prototype Selection Research Articles

Selection of effective training instances for scalable automatic image annotation

Automatic image annotation means employing learning models for describing visual contents of images by using text descriptors. With the fast growth of digital images in the web, large-scale automatic image annotation has started to deal with major challenges. The most important challenges are scalability and annotation performance. In this research, in order to solve scalability and the image annotation time challenge, the prototype selection approach is used. The assumption of the prototype selection is based on single-label instances while, in image annotation, an instance has more than one label. It means that instances are multi-label. Hence, to employ prototype selection algorithms in image annotation, focusing on the concept of multi-label is a critical task. Thus, taking an appropriate measure in these methods to compute the rate of dissimilarity between label vectors has a great importance. The proposed approach in this paper is based on multi-labeling of prototype selection methods by selecting a modifying appropriate binary dissimilarity measure, in comparison two label vectors. The effectiveness of the proposed approach in reducing the number of training instances and selecting effective ones has been shown by experiments on large-scale NUS-WIDE family image sets. The experimental results showed the effectiveness of the proposed approach in reducing the number of instances and improving annotation performance.

On the suitability of Prototype Selection methods for kNN classification with distributed data

In the current Information Age, data production and processing demands are ever increasing. This has motivated the appearance of large-scale distributed information. This phenomenon also applies to Pattern Recognition so that classic and common algorithms, such as the k-Nearest Neighbour, are unable to be used. To improve the efficiency of this classifier, Prototype Selection (PS) strategies can be used. Nevertheless, current PS algorithms were not designed to deal with distributed data, and their performance is therefore unknown under these conditions. This work is devoted to carrying out an experimental study on a simulated framework in which PS strategies can be compared under classical conditions as well as those expected in distributed scenarios. Our results report a general behaviour that is degraded as conditions approach to more realistic scenarios. However, our experiments also show that some methods are able to achieve a fairly similar performance to that of the non-distributed scenario. Thus, although there is a clear need for developing specific PS methodologies and algorithms for tackling these situations, those that reported a higher robustness against such conditions may be good candidates from which to start.

An experimental study on rank methods for prototype selection

Prototype selection is one of the most popular approaches for addressing the low efficiency issue typically found in the well-known k-Nearest Neighbour classification rule. These techniques select a representative subset from an original collection of prototypes with the premise of maintaining the same classification accuracy. Most recently, rank methods have been proposed as an alternative to develop new selection strategies. Following a certain heuristic, these methods sort the elements of the initial collection according to their relevance and then select the best possible subset by means of a parameter representing the amount of data to maintain. Due to the relative novelty of these methods, their performance and competitiveness against other strategies is still unclear. This work performs an exhaustive experimental study of such methods for prototype selection. A representative collection of both classic and sophisticated algorithms are compared to the aforementioned techniques in a number of datasets, including different levels of induced noise. Results report the remarkable competitiveness of these rank methods as well as their excellent trade-off between prototype reduction and achieved accuracy.

Improving nearest neighbor classification using Ensembles of Evolutionary Generated Prototype Subsets

One of the most accurate types of prototype selection algorithms, preprocessing techniques that select a subset of instances from the data before applying nearest neighbor classification to it, are evolutionary approaches. These algorithms result in very high accuracy and reduction rates, but unfortunately come at a substantial computational cost. In this paper, we introduce a framework that allows to efficiently use the intermediary results of the prototype selection algorithms to further increase their accuracy performance. Instead of only using the fittest prototype subset generated by the evolutionary algorithm, we use multiple prototype subsets in an ensemble setting. Secondly, in order to classify a test instance, we only use prototype subsets that accurately classify training instances in the neighborhood of that test instance. In an experimental evaluation, we apply our new framework to four state-of-the-art prototype selection algorithms and show that, by using our framework, more accurate results are obtained after less evaluations of the prototype selection method. We also present a case study with a prototype generation algorithm, showing that our framework is easily extended to other preprocessing paradigms as well.

Prototype generation on structural data using dissimilarity space representation

Data reduction techniques play a key role in instance-based classification to lower the amount of data to be processed. Among the different existing approaches, prototype selection (PS) and prototype generation (PG) are the most representative ones. These two families differ in the way the reduced set is obtained from the initial one: While the former aims at selecting the most representative elements from the set, the latter creates new data out of it. Although PG is considered to delimit more efficiently decision boundaries, the operations required are not so well defined in scenarios involving structural data such as strings, trees, or graphs. This work studies the possibility of using dissimilarity space (DS) methods as an intermediate process for mapping the initial structural representation to a statistical one, thereby allowing the use of PG methods. A comparative experiment over string data is carried out in which our proposal is faced to PS methods on the original space. Results show that the proposed strategy is able to achieve significantly similar results to PS in the initial space, thus standing as a clear alternative to the classic approach, with some additional advantages derived from the DS representation.

Optimal prototype selection for speech emotion recognition using fuzzy k-important nearest neighbour

Speech emotion recognition has been a popular topic of affective computing. Accuracy in speech emotion recognition depends on selecting the optimal prototype. In this paper, a new 2-D emotional speech recognition model based on a fuzzy k-important nearest neighbour FKINN and neuro-fuzzy network is described. In the FKINN algorithm, an important nearest neighbour selection rule is introduced. The neuro-fuzzy network applies a bounded sum of weighted fuzzy membership functions BSWFM. During the training process, BSWFM calculates the Takagi-Sugeno defuzzification values for the 2-D visual model. The emotional speech signals used in this work were obtained from the Berlin emotional speech database. The proposed new model achieves 83.5% overall classification accuracy with the 2-D emotional speech recognition model. The classification accuracies of anger, happiness, sadness, and neutral were 94.1%, 65.9%, 81.1%, and 87.5%, respectively.

Correcting the hub occurrence prediction bias in many dimensions

Data reduction is a common pre-processing step for k-nearest neighbor classification (kNN). The existing prototype selection methods implement different criteria for selecting relevant points to use in classification, which constitutes a selection bias. This study examines the nature of the instance selection bias in intrinsically high-dimensional data. In high-dimensional feature spaces, hubs are known to emerge as centers of influence in kNN classification. These points dominate most kNN sets and are often detrimental to classification performance. Our experiments reveal that different instance selection strategies bias the predictions of the behavior of hub-points in high-dimensional data in different ways. We propose to introduce an intermediate un-biasing step when training the neighbor occurrence models and we demonstrate promising improvements in various hubness-aware classification methods, on a wide selection of high-dimensional synthetic and real-world datasets.

Sparse regression and marginal testing using cluster prototypes

We propose a new approach for sparse regression and marginal testing, for data with correlated features. Our procedure first clusters the features, and then chooses as the cluster prototype the most informative feature in that cluster. Then we apply either sparse regression (lasso) or marginal significance testing to these prototypes. While this kind of strategy is not entirely new, a key feature of our proposal is its use of the post-selection inference theory of Taylor and others (2014, Exact post-selection inference for forward stepwise and least angle regression, Preprint, arXiv:1401.3889) and Lee and others (2014, Exact post-selection inference with the lasso, Preprint, arXiv:1311.6238v5) to compute exact [Formula: see text]-values and confidence intervals that properly account for the selection of prototypes. We also apply the recent "knockoff" idea of Barber and Candès (2014, Controlling the false discovery rate via knockoffs, Preprint, arXiv:1404.5609) to provide exact finite sample control of the FDR of our regression procedure. We illustrate our proposals on both real and simulated data.

최근접 이웃 규칙 기반 프로토타입 선택과 편의-분산을 이용한 성능 평가

최근접 이웃 규칙 기반 프로토타입 선택과 편의-분산을 이용한 성능 평가

SubMIL: Discriminative subspaces for multi-instance learning

As an important learning scheme for Multi-Instance Learning (MIL), the Instance Prototype (IP) selection-based MIL algorithms transform bags into a new instance feature space and achieve impressed classification performance. However, the number of IPs in the existing algorithms linearly increases with the scale of the training data. The performance and efficiencies of these algorithms are easily limited by the high dimension and noise when facing a large scale of training data. This paper proposes a discriminative subspaces-based instance prototype selection method that is suitable for reducing the computation complexity for large scale training data. In the proposed algorithm, we introduce the low-rank matrix recovery technique to find two discriminative and clean subspaces with less noise; then present a ℓ2,1 norm-based self-expressive sparse coding model to select the most representative instances in each subspace. Experimental results on several data sets show that our algorithm achieves superior and stable performance but with lower dimension compared with other IP selection strategies.

Вычислительная сложность отбора объектов и признаков для задач классификации с ограничениями монотонности

Decision rules with monotonicity constraints are often used in biomedical diagnostics. Simultaneous feature selection and prototype selection can significantly affect the degree of monotonicity of the data set and, as a consequence, the classification quality. In this paper we propose a systematization of discrete optimization problems of simultaneous feature selection and prototype selection and estimate their computational complexity.

Selection of virtual and physical prototypes in the product development process

The use of prototypes in the product development process (PDP) brings maturity to the project, reduces uncertainties, and helps in the conservation of information flow during the whole PDP. Usually, virtual prototypes (VP) are created in CAD/CAE systems, while physical prototypes (PP) are usually manufactured from techniques of material addition. In this reality, the creation/manufacturing of a prototype capable to meet its purpose will depend on the alternative available for its creation/manufacturing, and on the experience of the professional involved (designer). This work proposes a procedure to help in the selection of the best creation/manufacture technique for VP and PP, initially focusing on the automotive sector. The procedure was developed from theoretical backgrounds, discussion and consideration among professionals and scholars. The procedure was evaluated in a field research, where 34 professionals from 23 different companies of the automotive sector had been consulted. The theoretical frameworks, the proposed procedure, and the results of the field research are presented. Finally, the work is finished showing the conclusions and final considerations. The use of this procedure is expected to help experienced professionals and beginners, as well as scholars, in their research, education, and extension activities.

Novice Assistance Tool and Methodology: Design Decision and Task-Pattern Mapping

We present a possible method and tool to supportthe novice designers in the design tasks. Although in literaturestated that practitioners depend on memory and experiences,novice designers would take time to learn to acquire sufficient experience. The purpose of this research is to help novicedesigners to fulfill the usability goals that have determined by theexpert in designing a system. The proposed methodology and toolcan be considered as an assistance tool for novice designer indecision making in regards to prototype selection as well as thetask-pattern mapping template for pattern selection in fulfillingthe usability goal targeted. The tool was designed to manage thedecisions information and to recommend the chosen prototypebased on the Analytical Hierarchical Process (AHP) technique. Inoverall, about 87% of expert evaluators and prospective userswere agreed with the suggestion of prototype selection from theproposed tool. The findings from this study revealed that thecooperation from industry experts was valuable in determinedthe usability goal and direction and getting accumulativefeedbacks both from users and experts.

Decremental Sparse Modeling Representative Selection for prototype selection

Selecting few representatives or examples that can efficiently and reliably describe a set of data has always been a challenging task in computer vision and pattern recognition. Recently, Sparse Modeling Representative Selection (SMRS) was proposed as a powerful filter method for selecting the most relevant examples/instances in subspaces. The selection is achieved by ranking the examples using the L2 norm of the associated row in a coding matrix. This coding matrix is computed using data self-representativeness (the dictionary is given by the examples themselves) adopting block sparsity regularization. In this paper, we propose a decremental Sparse Modeling Representative Selection (D-SMRS) in which the selection of the representatives is broken down into several nested processes. The key contribution is a new scheme of sparse modeling that proceeds by progressive coding and pruning. It proceeds by eliminating outlier and noisy samples in the first stages so that the final stage (coding and selection) is performed on clean data. Thus, the final instance selection will not be heavily affected by the presence of outliers and aberrant samples. The proposed method was qualitatively and quantitatively evaluated. The qualitative evaluation concerned image selection and video summarization. The quantitative evaluation was performed on six benchmark image datasets using several state-of-the art selection methods with four different classifiers: 1-Nearest Neighbor (NN), Nearest Subspace (NS), Sparse Representation Classifier (SRC), and Support Vector Machines (SVM). The outlier rejection ability of the proposed method is also studied on real images. In all cases the selection computed by our algorithm achieved or outperformed existing state-of-the-art results.

Comparison of distributed evolutionary k-means clustering algorithms

Dealing with distributed data is one of the challenges for clustering, as most clustering techniques require the data to be centralized. One of them, k-means, has been elected as one of the most influential data mining algorithms for being simple, scalable, and easily modifiable to a variety of contexts and application domains. However, exact distributed versions of k-means are still sensitive to the selection of the initial cluster prototypes and require the number of clusters to be specified in advance. Additionally, preserving data privacy among repositories may be a complicating factor. In order to overcome k-means limitations, two different approaches were adopted in this paper: the first obtains a final model identical to the centralized version of the clustering algorithm and the second generates and selects clusters for each distributed data subset and combines them afterwards. It is also described how to apply the algorithms compared while preserving data privacy. The algorithms are compared experimentally from two perspectives: the theoretical one, through asymptotic complexity analyses, and the experimental one, through a comparative evaluation of results obtained from a collection of experiments and statistical tests. The results obtained indicate which algorithm is more suitable for each application scenario.

Prototype selection based on multi-objective optimisation and partition strategy

Prototype selection aims at reducing the storage of datasets and execution time, and improving prediction accuracy and operation efficiency by removing noisy or redundant samples, so the prediction classification accuracy is maximised and the reduction ratio is minimised simultaneously. To achieve this purpose, a two objective optimisation model is set up for prototype selection problem in the paper. To make the model be solved easier, it is transformed to a single objective optimisation model by the division of the two objectives, and a new two–layer genetic algorithm is proposed by using a divide–and–conquer partition strategy. The divide–and–conquer partition can divide the whole dataset into some random sub–datasets to be handled, respectively. The simulations are conducted and the proposed algorithm is compared with several existing algorithms. The results obtained on UCI static datasets and time series datasets indicate that the proposed algorithm is an expedient method in design nearest neighbour classifiers.

Two-Phase Usability Evaluation of Insurance Website Prototypes

This paper discusses the various challenges encountered during iterative testing of an insurance company's home page prototypes. The study focuses on the methodology details, including the selection of prototypes and usability evaluation methods and the considerations for practical trade-offs. During Phase 1, six individual think-aloud interviews were conducted to compare three prototypes. Participants were asked to complete tasks on the prototypes. Based on the findings from Phase 1, three prototypes were developed. In Phase 2, an online survey was administered with existing customers to compare these three prototypes. Survey responses indicated that content, layout, and visual appearance were most influential on users' preferences of the designs. At the end of our paper, we compare the implementation and the results of usability evaluation and the prototypes in the two phases and discuss the limitations of the study.

A Prototype Selection Method for Synthetic Object Recognition Based on Structural Graph Model

At present, many popular methods for object recognition are based on regional visual feature vectors which ignore the global structure of the object, leading to the “semantic gap” between image content and its tag. Graph-based structural representation of an object can record regional visual features and global relationships between regions that compose the object. However, the computation complexity of graph comparing limits the application of graph model in object recognition. One method for improving the recognizing speed is to reduce the prototypes in concept space of specific object category. We presented a prototype selection method for sythetic object recognition based on structural graph model using greedy clustering and representative sample selection. Experiments are conducted on a 2-D CAD synthetic object database. And the results show that the prototype selection method reduces the time cost of object recognition greatly with an acceptable loss in accuracy.

Data-Distribution-Aware Fuzzy Rough Set Model and its Application to Robust Classification.

Fuzzy rough sets (FRSs) are considered to be a powerful model for analyzing uncertainty in data. This model encapsulates two types of uncertainty: 1) fuzziness coming from the vagueness in human concept formation and 2) roughness rooted in the granulation coming with human cognition. The rough set theory has been widely applied to feature selection, attribute reduction, and classification. However, it is reported that the classical FRS model is sensitive to noisy information. To address this problem, several robust models have been developed in recent years. Nevertheless, these models do not consider a statistical distribution of data, which is an important type of uncertainty. Data distribution serves as crucial information for designing an optimal classification or regression model. Thus, we propose a data-distribution-aware FRS model that considers distribution information and incorporates it in computing lower and upper fuzzy approximations. The proposed model considers not only the similarity between samples, but also the probability density of classes. In order to demonstrate the effectiveness of the proposed model, we design a new sample evaluation index for prototype-based classification based on the model, and a prototype selection algorithm is developed using this index. Furthermore, a robust classification algorithm is constructed with prototype covering and nearest neighbor classification. Experimental results confirm the robustness and effectiveness of the proposed model.

Improving kNN multi-label classification in Prototype Selection scenarios using class proposals

Prototype Selection (PS) algorithms allow a faster Nearest Neighbor classification by keeping only the most profitable prototypes of the training set. In turn, these schemes typically lower the performance accuracy. In this work a new strategy for multi-label classifications tasks is proposed to solve this accuracy drop without the need of using all the training set. For that, given a new instance, the PS algorithm is used as a fast recommender system which retrieves the most likely classes. Then, the actual classification is performed only considering the prototypes from the initial training set belonging to the suggested classes. Results show that this strategy provides a large set of trade-off solutions which fills the gap between PS-based classification efficiency and conventional kNN accuracy. Furthermore, this scheme is not only able to, at best, reach the performance of conventional kNN with barely a third of distances computed, but it does also outperform the latter in noisy scenarios, proving to be a much more robust approach.