Abstract: This paper presents content-based image retrieval (CBIR) frameworks with relevance feedback (RF) based on combined learning of support vector machines (SVM) and AdaBoosts. The framework incorporates only most relevant images obtained from both the learning algorithm. To speed up the system, it removes irrelevant images from the database, which are returned from SVM learner. It is the key to achieve the effective retrieval performance in terms of time and accuracy. The experimental results show that this framework had significant improvement in retrieval effectiveness, which can finally improve the retrieval performance.
Abstract: Various cis-regulatory module (CRM) predictors have been proposed in the last decade. Several well-established CRM predictors adopted different categories of prediction strategies, including window clustering, probabilistic modeling and phylogenetic footprinting. Appropriate integration of them has a potential to achieve high quality CRM prediction. This study analyzed four existing CRM predictors (ClusterBuster, MSCAN, CisModule and MultiModule) to seek a predictor combination that delivers a higher accuracy than individual CRM predictors. 465 CRMs across 140 Drosophila melanogaster genes from the RED fly database were used to evaluate the integrated CRM predictor proposed in this study. The results show that four predictor combinations achieved superior performance than the best individual CRM predictor.
Abstract: The Choquet integral is a tool for the information fusion that is very effective in the case where fuzzy measures associated with it are well chosen. In this paper, we propose a new approach for calculating fuzzy measures associated with the Choquet integral in a context of data fusion in multimodal biometrics. The proposed approach is based on genetic algorithms. It has been validated in two databases: the first base is relative to synthetic scores and the second one is biometrically relating to the face, fingerprint and palmprint. The results achieved attest the robustness of the proposed approach.
Abstract: Multimodal biometric systems integrate the data presented by multiple biometric sources, hence offering a better performance than the systems based on a single biometric modality. Although the coupling of biometric systems can be done at different levels, the fusion at the scores level is the most common since it has been proven effective than the rest of the fusion levels. However, the scores from different modalities are generally heterogeneous. A step of normalizing the scores is needed to transform these scores into a common domain before combining them. In this paper, we study the performance of several normalization techniques with various fusion methods in a context relating to the merger of three unimodal systems based on the face, the palmprint and the fingerprint. We also propose a new adaptive normalization method that takes into account the distribution of client scores and impostor scores. Experiments conducted on a database of 100 people show that the performances of a multimodal system depend on the choice of the normalization method and the fusion technique. The proposed normalization method has given the best results.
Abstract: In this paper, a study of slope failures along the Alishan Highway is carried out. An innovative empirical model is developed based on 15-year records of rainfall-induced slope failures. The statistical models are intended for assessing the volume of landslide for slope failure along the Alishan Highway in the future. The rainfall data considered in the proposed models include the effective cumulative rainfall and the critical rainfall intensity. The effective cumulative rainfall is defined at the point when the curve of cumulative rainfall goes from steep to flat. Then, the rainfall thresholds of landslide are established for assessing the volume of landslide and issuing warning and/or closure for the Alishan Highway during a future extreme rainfall. Slope failures during Typhoon Saola in 2012 demonstrate that the new empirical model is effective and applicable to other cases with similar rainfall conditions.
Abstract: Object identification and segmentation application requires extraction of object in foreground from the background. In this paper the Bhattacharya distance based probabilistic approach is utilized with an active contour model (ACM) to segment an object from the background. In the proposed approach, the Bhattacharya histogram is calculated on non-linear structure tensor space. Based on the histogram, new formulation of active contour model is proposed to segment images. The results are tested on both color and gray images from the Berkeley image database. The experimental results show that the proposed model is applicable to both color and gray images as well as both texture images and natural images. Again in comparing to the Bhattacharya based ACM in ICA space, the proposed model is able to segment multiple object too.
Abstract: Augmented Reality is an application which combines a live view of real-world environment and computer-generated images. This paper studies and demonstrates an efficient Augmented Reality development in the mobile Android environment with the native Java language and Android SDK. Major components include Barcode Reader, File Loader, Marker Detector, Transform Matrix Generator, and a cloud database.
Abstract: Industrial robotic arms utilize multiple end-effectors, each for a specific part and for a specific task. We propose a novel algorithm which will define a single end-effector’s configuration able to grasp a given set of objects with different geometries. The algorithm will have great benefit in production lines allowing a single robot to grasp various parts. Hence, reducing the number of endeffectors needed. Moreover, the algorithm will reduce end-effector design and manufacturing time and final product cost. The algorithm searches for a common grasp over the set of objects. The search algorithm maps all possible grasps for each object which satisfy a quality criterion and takes into account possible external wrenches (forces and torques) applied to the object. The mapped grasps are- represented by high-dimensional feature vectors which describes the shape of the gripper. We generate a database of all possible grasps for each object in the feature space. Then we use a search and
classification algorithm for intersecting all possible grasps over all
parts and finding a single common grasp suitable for all objects.
We present simulations of planar and spatial objects to validate the
feasibility of the approach.
Abstract: The paper discusses the implementation of the MultiAgent classification System (MACS) and utilizing it to provide an automated and accurate classification of end users developing applications in the spreadsheet domain. However, different technologies have been brought together to build MACS. The strength of the system is the integration of the agent technology with the FIPA specifications together with other technologies, which are the .NET widows service based agents, the Windows Communication Foundation (WCF) services, the Service Oriented Architecture (SOA), and Oracle Data Mining (ODM). The Microsoft's .NET widows service based agents were utilized to develop the monitoring agents of MACS, the .NET WCF services together with SOA approach allowed the distribution and communication between agents over the WWW. The Monitoring Agents (MAs) were configured to execute automatically to monitor excel spreadsheets development activities by content. Data gathered by the Monitoring Agents from various resources over a period of time was collected and filtered by a Database Updater Agent (DUA) residing in the .NET client application of the system. This agent then transfers and stores the data in Oracle server database via Oracle stored procedures for further processing that leads to the classification of the end user developers.
Abstract:
ankings for output of Chinese main agricultural commodity in the world for 1978, 1980, 1990, 2000, 2006, 2007 and 2008 have been released in United Nations FAO Database. Unfortunately, where the ranking of output of Chinese cotton lint in the world for 2008 was missed. This paper uses sequential data mining methods with decision rules filling this gap. This new data mining method will be help to give a further improvement for United Nations FAO Database.
Abstract: Free access for Georgian goods to the EU markets is one of the important factors for Georgia’s economic development, attraction of investments and raising the standard of living. The European Union is the most important trade partner for Georgia. Great experience has been accumulated with respect to removing trade barriers between Georgia and the European Union. Despite it, certain problems still persist.
In the present article, we have reviewed the systems of preferences with the European Union, the EU’s Generalized System of Preferences (GSP) and the essence of ongoing reform; we have assessed weak and strong sides of relations established between the European Union and Georgia in this regard; analyzed Georgia’s export and import over the past years; also reviewed the prospect of a unified database; established existing and anticipated positive and negative factors. Based on the analysis, we have provided the relevant recommendations.
Abstract: Standalone micro-hydrokinetic river (MHR) system is
one of the promising technologies to be used for remote rural
electrification. It simply requires the flow of water instead of
elevation or head, leading to expensive civil works. This paper
demonstrates an economic benefit offered by a standalone MHR
system when compared to the commonly used standalone systems
such as solar, wind and diesel generator (DG) at the selected study
site in Kwazulu Natal. Wind speed and solar radiation data of the
selected rural site have been taken from national aeronautics and
space administration (NASA) surface meteorology database. The
hybrid optimization model for electric renewable (HOMER) software
was used to determine the most feasible solution when using MHR,
solar, wind or DG system to supply 5 rural houses. MHR system
proved to be the best cost-effective option to consider at the study site
due to its low cost of energy (COE) and low net present cost (NPC).
Abstract: Tanzania Secondary Schools e-Learning (TanSSe-L) system is a customized learning management system (LMS) developed to enable ICT support in teaching and learning functions. Methodologies involved in the development of TanSSe-L system are Object oriented system analysis and design with UML to create and model TanSSe-L system database structure in the form of a design class diagram, Model Driven Architecture (MDA) to provide a well defined process in TanSSe-L system development, where MDA conceptual layers were integrated with system development life cycle and customization of open source learning management system which was used during implementation stage to create a timely functional TanSSe-L system. Before customization, a base for customization was prepared. This was the manual transformation from TanSSe-L system platform independent models (PIM) to TanSSe-L system specific PIM. This paper presents how Moodle open source LMS was analyzed and prepared to be the TanSSe-L system specific PIM as applied by MDA.
Abstract: This paper explains how the New Jersey Institute of Technology surveying student team members designed and created an interactive GIS map, the purpose of which is to be useful to the land surveyor and engineer for project management. This was achieved by building a research and storage database that can be easily integrated into any land surveyor’s current operations through the use of ArcGIS 10, Arc Catalog, and AutoCAD. This GIS database allows for visual representation and information querying for multiple job sites, and simple access to uploaded data, which is geospatially referenced to each individual job site or project. It can also be utilized by engineers to determine design criteria, or to store important files. This cost-effective approach to a surveying map not only saves time, but saves physical storage space and paper resources.
Abstract: In this paper various techniques in relation to large-scale systems are presented. At first, explanation of large-scale systems and differences from traditional systems are given. Next, possible specifications and requirements on hardware and software are listed. Finally, examples of large-scale systems are presented.
Abstract: The exponential increase in the volume of medical image database has imposed new challenges to clinical routine in maintaining patient history, diagnosis, treatment and monitoring. With the advent of data mining and machine learning techniques it is possible to automate and/or assist physicians in clinical diagnosis. In this research a medical image classification framework using data mining techniques is proposed. It involves feature extraction, feature selection, feature discretization and classification. In the classification phase, the performance of the traditional kNN k nearest neighbor classifier is improved using a feature weighting scheme and a distance weighted voting instead of simple majority voting. Feature weights are calculated using the interestingness measures used in association rule mining. Experiments on the retinal fundus images show that the proposed framework improves the classification accuracy of traditional kNN from 78.57 % to 92.85 %.
Abstract: In order to integrate knowledge in heterogeneous
case-based reasoning (CBR) systems, ontology-based CBR system
has become a hot topic. To solve the facing problems of
ontology-based CBR system, for example, its architecture is
nonstandard, reusing knowledge in legacy CBR is deficient, ontology
construction is difficult, etc, we propose a novel approach for
semi-automatically construct ontology-based CBR system whose
architecture is based on two-layer ontology. Domain knowledge
implied in legacy case bases can be mapped from relational database
schema and knowledge items to relevant OWL local ontology
automatically by a mapping algorithm with low time-complexity. By
concept clustering based on formal concept analysis, computing
concept equation measure and concept inclusion measure, some
suggestions about enriching or amending concept hierarchy of OWL
local ontologies are made automatically that can aid designers to
achieve semi-automatic construction of OWL domain ontology.
Validation of the approach is done by an application example.
Abstract: In this paper we introduce an effective ECG compression algorithm based on two dimensional multiwavelet transform. Multiwavelets offer simultaneous orthogonality, symmetry and short support, which is not possible with scalar two-channel wavelet systems. These features are known to be important in signal processing. Thus multiwavelet offers the possibility of superior performance for image processing applications. The SPIHT algorithm has achieved notable success in still image coding. We suggested applying SPIHT algorithm to 2-D multiwavelet transform of2-D arranged ECG signals. Experiments on selected records of ECG from MIT-BIH arrhythmia database revealed that the proposed algorithm is significantly more efficient in comparison with previously proposed ECG compression schemes.
Abstract: This paper proposes a stroke extraction method for use in off-line signature verification. After giving a brief overview of the current ongoing researches an algorithm is introduced for detecting and following strokes in static images of signatures. Problems like the handling of junctions and variations in line width and line intensity are discussed in detail. Results are validated by both using an existing on-line signature database and by employing image registration methods.
Abstract: In Content-Based Image Retrieval systems it is
important to use an efficient indexing technique in order to perform
and accelerate the search in huge databases. The used indexing
technique should also support the high dimensions of image features.
In this paper we present the hierarchical index NOHIS-tree (Non
Overlapping Hierarchical Index Structure) when we scale up to very
large databases. We also present a study of the influence of clustering
on search time. The performance test results show that NOHIS-tree
performs better than SR-tree. Tests also show that NOHIS-tree keeps
its performances in high dimensional spaces. We include the
performance test that try to determine the number of clusters in
NOHIS-tree to have the best search time.