Incremental Linear Discriminant Analysis Using Sufficient Spanning Sets and Its Applications

This paper presents an incremental learning solution for Linear Discriminant Analysis (LDA) and its applications to object recognition problems. We apply the sufficient spanning set approximation in three steps i.e. update for the total scatter matrix, between-class scatter matrix and the projected data matrix, which leads an online solution which closely agrees with the batch solution in accuracy while significantly reducing the computational complexity. The algorithm yields an efficient solution to incremental LDA even when the number of classes as well as the set size is large. The incremental LDA method has been also shown useful for semi-supervised online learning. Label propagation is done by integrating the incremental LDA into an EM framework. The method has been demonstrated in the task of merging large datasets which were collected during MPEG standardization for face image retrieval, face authentication using the BANCA dataset, and object categorisation using the Caltech101 dataset.     

Region-based parallelization of irregular reductions on explicitly managed memory hierarchies

Multicore architectures are evolving with the promise of extreme performance for the classes of applications that require high performance and large bandwidth of memory. Irregular reduction is one of important computation patterns for many complex scientific applications, and it typically requires high performance and large bandwidth of memory. In this article, we propose region-based parallelization techniques for irregular reductions on multicore architectures with explicitly managed memory hierarchies. Managing memory hierarchy in software requires a lot of programming efforts and tends to be error-prone. The difficulties are even worse for applications with irregular data access patterns. To relieve the burden of memory management from programmers, we develop abstractions, particularly targeted to irregular reduction, for structuring parallel tasks, mapping the parallel tasks to processing units and scheduling data transfers between the memory hierarchies. Our framework employs iteration reordering based on regions of data along with dynamic scheduling of parallel tasks. We experimentally evaluate the effectiveness of our techniques for irregular reduction kernels on the Cell processor embedded in a Sony PlayStation3. Experimental results show the speedups of 8 to 14 on the six available SPEs.     

Plasmonic enhanced fluorescence spectroscopy using side-polished microstructured optical fiber

We report excitation of surface plasmon in a gold-coated side-polished D-shape microstructure optical fiber (MOF). As the leaky evanescent field from the fiber core becomes highly localized by the plasmon wave, its intensity also gets amplified significantly. Here we demonstrate an efficient use of this intensified field as excitation in fluorescence spectroscopy. The so-called plasmonic enhanced fluorescence emission from Rhodamine B has been investigated experimentally. First, plasmonic effect alone was found to provide an immediate fluorescence enhancement factor of two. Second, experimental results show a good agreement with theoretical modeling. Strong evanescent field generation and surface enhancement with simple metallic coating makes this fiber based device a good candidate for compact fluorescence spectroscopy.     

Collaborative user modeling with user-generated tags for social recommender systems

With the popularity of social media services, the sheer amount of content is increasing exponentially on the Social Web that leads to attract considerable attention to recommender systems. Recommender systems provide users with recommendations of items suited to their needs. To provide proper recommendations to users, recommender systems require an accurate user model that can reflect a user’s characteristics, preferences and needs. In this study, by leveraging user-generated tags as preference indicators, we propose a new collaborative approach to user modeling that can be exploited to recommender systems. Our approach first discovers relevant and irrelevant topics for users, and then enriches an individual user model with collaboration from other similar users. In order to evaluate the performance of our model, we compare experimental results with a user model based on collaborative filtering approaches and a vector space model. The experimental results have shown the proposed model provides a better representation in user interests and achieves better recommendation results in terms of accuracy and ranking.     

A fact-oriented ontological approach to human process modeling for knowledge-intensive business services

Modern business environment emphasizes the role of knowledge-intensive business services (KIBS). As well as the enterprises in any other kinds of industries, business process management (BPM) can be a source of core competency for the enterprises in KIBS industry. However, most of the business processes in KIBS are human processes which are collaborative, innovative, and dynamic, which cannot be supported well by current BPM technologies. Human interaction management (HIM) has emerged as an alternative theory to deal with human processes in KIBS. But, the current ways of process modeling and management in HIM have a problem of complexity, especially in terms of realizing its principle of ‘supportive activity management’. This research adopts fact-oriented ontological approach to address this problem. Based on fact-oriented approach, human-friendly condition modeling and state management during supportive activity management can be achieved.     

A quay crane scheduling algorithm considering the workload of yard cranes in a container yard

Quay cranes (QCs) are the most important equipment for vessel operation in container terminals. It is important to efficiently schedule the operations of QCs so that a high productivity may be achieved in terminals. Because the storage yard is shared by multiple vessels, where the congestion of yard cranes and yard trucks must be considered, the operations of QCs for multiple vessels must be simultaneously considered for QC scheduling. The objective terms include the QC make-span, the total traveling distance of QCs, the expected delay time from interference, and the uniformity of workload among blocks. This study suggests a heuristic algorithm based on a meta-heuristic search algorithm. Finally, a numerical experiment is carried out to test the performance of the algorithm.     

Individual,social, and organizational contexts for active knowledge sharing in communities of practice

Firms that have implemented knowledge management initiatives are now interested in nurturing voluntary knowledge sharing organizations, called communities of practice (CoPs). Adopting the Triandis model on attitude formation, we identified and validated a set of organizational factors that was anticipated to have effects on knowledge sharing by CoP members such as perceived consequences, affect, social factors and facilitating conditions. One hundred and seventy-nine members from 70 CoPs of a large multinational electronics firm participated in this survey. Based on the PLS analysis, perceived consequences, affect, social factors, and facilitating conditions were found to significantly affect knowledge sharing in CoPs.     

A new weighted approach to imbalanced data classification problem via support vector machine with quadratic cost function

In this paper, a new weighted approach on Lagrangian support vector machine for imbalanced data classification problem is proposed. The weight parameters are embedded in the Lagrangian SVM formulation. The training method for weighted Lagrangian SVM is presented and its convergence is proven. The weighted Lagrangian SVM classifier is tested and compared with some other SVMs using synthetic and real data to show its effectiveness and feasibility.     

Speed scaling with a solar cell

We consider the setting of a device that obtains its energy from a battery and some regenerative source such as a solar cell. We consider the speed scaling problem of scheduling a collection of tasks with release times, deadlines, and sizes, so as to minimize the energy recharge rate of the regenerative source. This is the first theoretical investigation of speed scaling for devices with a regenerative energy source. We show that the problem can be expressed as a polynomial sized convex program. We show that, using the KKT conditions, one can obtain an efficient algorithm to verify the optimality of a schedule. We show that the energy optimal YDS schedule is 2-approximate with respect to the recharge rate. We show that the online algorithm BKP is O(1)$O\left(1\right)$-competitive with respect to recharge rate.