Search results for: Kekre's Fast Search

Authors: Dr. H. B. Kekre, Sudeep D. Thepade

Abstract:

Panoramic view generation has always offered novel and distinct challenges in the field of image processing. Panoramic view generation is nothing but construction of bigger view mosaic image from set of partial images of the desired view. The paper presents a solution to one of the problems of image seascape formation where some of the partial images are color and others are grayscale. The simplest solution could be to convert all image parts into grayscale images and fusing them to get grayscale image panorama. But in the multihued world, obtaining the colored seascape will always be preferred. This could be achieved by picking colors from the color parts and squirting them in grayscale parts of the seascape. So firstly the grayscale image parts should be colored with help of color image parts and then these parts should be fused to construct the seascape image. The problem of coloring grayscale images has no exact solution. In the proposed technique of panoramic view generation, the job of transferring color traits from reference color image to grayscale image is done by palette based method. In this technique, the color palette is prepared using pixel windows of some degrees taken from color image parts. Then the grayscale image part is divided into pixel windows with same degrees. For every window of grayscale image part the palette is searched and equivalent color values are found, which could be used to color grayscale window. For palette preparation we have used RGB color space and Kekre-s LUV color space. Kekre-s LUV color space gives better quality of coloring. The searching time through color palette is improved over the exhaustive search using Kekre-s fast search technique. After coloring the grayscale image pieces the next job is fusion of all these pieces to obtain panoramic view. For similarity estimation between partial images correlation coefficient is used.

Keywords: Panoramic View, Similarity Estimate, Color Transfer, Color Palette, Kekre's Fast Search, Kekre's LUV

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1753

Authors: H. B. Kekre, Tanuja K. Sarode, Bhakti Raul

Abstract:

In this paper we propose segmentation approach based on Vector Quantization technique. Here we have used Kekre-s fast codebook generation algorithm for segmenting low-altitude aerial image. This is used as a preprocessing step to form segmented homogeneous regions. Further to merge adjacent regions color similarity and volume difference criteria is used. Experiments performed with real aerial images of varied nature demonstrate that this approach does not result in over segmentation or under segmentation. The vector quantization seems to give far better results as compared to conventional on-the-fly watershed algorithm.

Keywords: Image Segmentation, , Codebook, Codevector, data compression, Encoding

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2195

Authors: H. B. Kekre, Tanuja K. Sarode

Abstract:

This paper presents a very simple and efficient algorithm for codebook search, which reduces a great deal of computation as compared to the full codebook search. The algorithm is based on sorting and centroid technique for search. The results table shows the effectiveness of the proposed algorithm in terms of computational complexity. In this paper we also introduce a new performance parameter named as Average fractional change in pixel value as we feel that it gives better understanding of the closeness of the image since it is related to the perception. This new performance parameter takes into consideration the average fractional change in each pixel value.

Keywords: Vector Quantization, Data Compression, Encoding, Searching.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1613

Authors: Lai-Man Po, Chi-Wang Ting, Ka-Ho Ng

Abstract:

Recently, an enhanced hexagon-based search (EHS) algorithm was proposed to speedup the original hexagon-based search (HS) by exploiting the group-distortion information of some evaluated points. In this paper, a second version of the EHS is proposed with a new point-oriented inner search technique which can further speedup the HS in both large and small motion environments. Experimental results show that the enhanced hexagon-based search version-2 (EHS2) is faster than the HS up to 34% with negligible PSNR degradation.

Keywords: Inner search, fast motion estimation, block-matching, hexagon search

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1432

Authors: Feifei Lee, Qiu Chen, Koji Kotani, Tadahiro Ohmi

Abstract:

In this paper, we propose a novel fast search algorithm for short MPEG video clips from video database. This algorithm is based on the adjacent pixel intensity difference quantization (APIDQ) algorithm, which had been reliably applied to human face recognition previously. An APIDQ histogram is utilized as the feature vector of the frame image. Instead of fully decompressed video frames, partially decoded data, namely DC images are utilized. Combined with active search [4], a temporal pruning algorithm, fast and robust video search can be realized. The proposed search algorithm has been evaluated by 6 hours of video to search for given 200 MPEG video clips which each length is 15 seconds. Experimental results show the proposed algorithm can detect the similar video clip in merely 80ms, and Equal Error Rate (ERR) of 3 % is achieved, which is more accurately and robust than conventional fast video search algorithm.

Keywords: Fast search, adjacent pixel intensity difference quantization (APIDQ), DC image, histogram feature.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1580

Authors: Qiu Chen, Feifei Lee, Koji Kotani, Tadahiro Ohmi

Abstract:

In this paper, we propose an efficient hierarchical DNA sequence search method to improve the search speed while the accuracy is being kept constant. For a given query DNA sequence, firstly, a fast local search method using histogram features is used as a filtering mechanism before scanning the sequences in the database. An overlapping processing is newly added to improve the robustness of the algorithm. A large number of DNA sequences with low similarity will be excluded for latter searching. The Smith-Waterman algorithm is then applied to each remainder sequences. Experimental results using GenBank sequence data show the proposed method combining histogram information and Smith-Waterman algorithm is more efficient for DNA sequence search.

Keywords: Fast search, DNA sequence, Histogram feature, Smith-Waterman algorithm, Local search

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1330

Authors: Feifei Lee, Qiu Chen, Koji Kotani, Tadahiro Ohmi

Abstract:

In this paper, we propose an improved fast search algorithm using combined histogram features and temporal division method for short MPEG video clips from large video database. There are two types of histogram features used to generate more robust features. The first one is based on the adjacent pixel intensity difference quantization (APIDQ) algorithm, which had been reliably applied to human face recognition previously. An APIDQ histogram is utilized as the feature vector of the frame image. Another one is ordinal feature which is robust to color distortion. Combined with active search [4], a temporal pruning algorithm, fast and robust video search can be realized. The proposed search algorithm has been evaluated by 6 hours of video to search for given 200 MPEG video clips which each length is 30 seconds. Experimental results show the proposed algorithm can detect the similar video clip in merely 120ms, and Equal Error Rate (ERR) of 1% is achieved, which is more accurately and robust than conventional fast video search algorithm.

Keywords: Fast search, Adjacent pixel intensity differencequantization (APIDQ), DC image, Histogram feature.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1625

Authors: T. Song, T. Shimamoto

Abstract:

In this paper a fast motion estimation method for H.264/AVC named Triplet Search Motion Estimation (TS-ME) is proposed. Similar to some of the traditional fast motion estimation methods and their improved proposals which restrict the search points only to some selected candidates to decrease the computation complexity, proposed algorithm separate the motion search process to several steps but with some new features. First, proposed algorithm try to search the real motion area using proposed triplet patterns instead of some selected search points to avoid dropping into the local minimum. Then, in the localized motion area a novel 3-step motion search algorithm is performed. Proposed search patterns are categorized into three rings on the basis of the distance from the search center. These three rings are adaptively selected by referencing the surrounding motion vectors to early terminate the motion search process. On the other hand, computation reduction for sub pixel motion search is also discussed considering the appearance probability of the sub pixel motion vector. From the simulation results, motion estimation speed improved by a factor of up to 38 when using proposed algorithm than that of the reference software of H.264/AVC with ignorable picture quality loss.

Keywords: Motion estimation, VLSI, image processing, search patterns

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1333

Authors: Feifei Lee, Qiu Chen, Koji Kotani, Tadahiro Ohmi

Abstract:

In this paper, we present an improved fast and robust search algorithm for copy detection using histogram-based features for short MPEG video clips from large video database. There are two types of histogram features used to generate more robust features. The first one is based on the adjacent pixel intensity difference quantization (APIDQ) algorithm, which had been reliably applied to human face recognition previously. An APIDQ histogram is utilized as the feature vector of the frame image. Another one is ordinal histogram feature which is robust to color distortion. Furthermore, by Combining with a temporal division method, the spatial and temporal features of the video sequence are integrated to realize fast and robust video search for copy detection. Experimental results show the proposed algorithm can detect the similar video clip more accurately and robust than conventional fast video search algorithm.

Keywords: Fast search, Copy detection, Adjacent pixel intensity difference quantization (APIDQ), DC image, Histogram feature.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1450

Authors: Hyungkeun Jee, Jooyoung Lee, Dowon Hong

Abstract:

The most common forensic activity is searching a hard disk for string of data. Nowadays, investigators and analysts are increasingly experiencing large, even terabyte sized data sets when conducting digital investigations. Therefore consecutive searching can take weeks to complete successfully. There are two primary search methods: index-based search and bitwise search. Index-based searching is very fast after the initial indexing but initial indexing takes a long time. In this paper, we discuss a high speed bitwise search model for large-scale digital forensic investigations. We used pattern matching board, which is generally used for network security, to search for string and complex regular expressions. Our results indicate that in many cases, the use of pattern matching board can substantially increase the performance of digital forensic search tools.

Keywords: Digital forensics, search, regular expression.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1807

Authors: Trung Hieu Tran, Hyo-Moon Cho, Sang-Bock Cho

Abstract:

Motion estimation is the most computationally intensive part in video processing. Many fast motion estimation algorithms have been proposed to decrease the computational complexity by reducing the number of candidate motion vectors. However, these studies are for fast search algorithms themselves while almost image and video compressions are operated with software based. Therefore, the timing constraints for running these motion estimation algorithms not only challenge for the video codec but also overwhelm for some of processors. In this paper, the performance of motion estimation is enhanced by using Intel's Streaming SIMD Extension 2 (SSE2) technology with Intel Pentium 4 processor.

Keywords: Motion Estimation, Full Search, Three StepSearch, MMX/SSE/SSE2 Technologies, SIMD.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2101

Authors: Karunakar A K, Manohara Pai M M

Abstract:

In 3D-wavelet video coding framework temporal filtering is done along the trajectory of motion using Motion Compensated Temporal Filtering (MCTF). Hence computationally efficient motion estimation technique is the need of MCTF. In this paper a predictive technique is proposed in order to reduce the computational complexity of the MCTF framework, by exploiting the high correlation among the frames in a Group Of Picture (GOP). The proposed technique applies coarse and fine searches of any fast block based motion estimation, only to the first pair of frames in a GOP. The generated motion vectors are supplied to the next consecutive frames, even to subsequent temporal levels and only fine search is carried out around those predicted motion vectors. Hence coarse search is skipped for all the motion estimation in a GOP except for the first pair of frames. The technique has been tested for different fast block based motion estimation algorithms over different standard test sequences using MC-EZBC, a state-of-the-art scalable video coder. The simulation result reveals substantial reduction (i.e. 20.75% to 38.24%) in the number of search points during motion estimation, without compromising the quality of the reconstructed video compared to non-predictive techniques. Since the motion vectors of all the pair of frames in a GOP except the first pair will have value ±1 around the motion vectors of the previous pair of frames, the number of bits required for motion vectors is also reduced by 50%.

Keywords: Motion Compensated Temporal Filtering, predictivemotion estimation, lifted wavelet transform, motion vector

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1619

Authors: O. Saravanan, A. Nagappan, P. Gnanasekar, S. Sharavanan, D. Vinodkumar, T. Elayabharathi, G. Karthik

Abstract:

The multi-agent system for processing Bio-signals will help the medical practitioners to have a standard examination procedure stored in web server. Web Servers supporting any standard Search Engine follow all possible combinations of the search keywords as an input by the user to a Search Engine. As a result, a huge number of Web-pages are shown in the Web browser. It also helps the medical practitioner to interact with the expert in the field his need in order to make a proper judgment in the diagnosis phase [3].A web server uses a web server plug in to establish and maintained the medical practitioner to make a fast analysis. If the user uses the web server client can get a related data requesting their search. DB agent, EEG / ECG / EMG agents- user placed with difficult aspects for updating medical information-s in web server.

Keywords: DB agent, EEG, ECG, EMG, Web server agent, JADE

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2076

Authors: Kim Nee Goh, Viknesh Kumar Naleyah

Abstract:

Online trading is an alternative to conventional shopping method. People trade goods which are new or pre-owned before. However, there are times when a user is not able to search the items wanted online. This is because the items may not be posted as yet, thus ending the search. Conventional search mechanism only works by searching and matching search criteria (requirement) with data available in a particular database. This research aims to match current search requirements with future postings. This would involve the time factor in the conventional search method. A Car Matching Alert System (CMAS) prototype was developed to test the matching algorithm. When a buyer-s search returns no result, the system saves the search and the buyer will be alerted if there is a match found based on future postings. The algorithm developed is useful and as it can be applied in other search context.

Keywords: Matching algorithm, online trading, search, future postings, car matching

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1427

Authors: Sushma Satpute, Vaishali Katkar, Nilesh Sahare

Abstract:

XML files contain data which is in well formatted manner. By studying the format or semantics of the grammar it will be helpful for fast retrieval of the data. There are many algorithms which describes about searching the data from XML files. There are no. of approaches which uses data structure or are related to the contents of the document. In these cases user must know about the structure of the document and information retrieval techniques using NLPs is related to content of the document. Hence the result may be irrelevant or not so successful and may take more time to search.. This paper presents fast XML retrieval techniques by using new indexing technique and the concept of RXML. When indexing an XML document, the system takes into account both the document content and the document structure and assigns the value to each tag from file. To query the system, a user is not constrained about fixed format of query.

Keywords: XML Retrieval, Indexed Search, Information Retrieval.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1783

Authors: Amol Bapuso Rajmane, Pradeep M. Patil, Prakash J. Kulkarni

Abstract:

The Information Retrieval community is facing the problem of effective representation of Web search results. When we organize web search results into clusters it becomes easy to the users to quickly browse through search results. The traditional search engines organize search results into clusters for ambiguous queries, representing each cluster for each meaning of the query. The clusters are obtained according to the topical similarity of the retrieved search results, but it is possible for results to be totally dissimilar and still correspond to the same meaning of the query. People search is also one of the most common tasks on the Web nowadays, but when a particular person’s name is queried the search engines return web pages which are related to different persons who have the same queried name. By placing the burden on the user of disambiguating and collecting pages relevant to a particular person, in this paper, we have developed an approach that clusters web pages based on the association of the web pages to the different people and clusters that are based on generic entity search.

Keywords: Entity resolution, information retrieval, graph based disambiguation, web people search, clustering.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1504

Authors: Linda, Sau-ling LAI

Abstract:

This paper compares the search engine marketing strategies adopted in China and the Western countries through two illustrative cases, namely, Google and Baidu. Marketers in the West use search engine optimization (SEO) to rank their sites higher for queries in Google. Baidu, however, offers paid search placement, or the selling of engine results for particular keywords to the higher bidders. Whereas Google has been providing innovative services ranging from Google Map to Google Blog, Baidu remains focused on search services – the one that it does best. The challenges and opportunities of the Chinese Internet market offered to global entrepreneurs are also discussed in the paper

Keywords: Search Engine, Web analytics, Google, Baidu

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2456

Authors: R. Lokeshkumar, P. Sengottuvelan

Abstract:

Web mining is to discover and extract useful Information. Different users may have different search goals when they search by giving queries and submitting it to a search engine. The inference and analysis of user search goals can be very useful for providing an experience result for a user search query. In this project, we propose a novel approach to infer user search goals by analyzing search web logs. First, we propose a novel approach to infer user search goals by analyzing search engine query logs, the feedback sessions are constructed from user click-through logs and it efficiently reflect the information needed for users. Second we propose a preprocessing technique to clean the unnecessary data’s from web log file (feedback session). Third we propose a technique to generate pseudo-documents to representation of feedback sessions for clustering. Finally we implement k-medoids clustering algorithm to discover different user search goals and to provide a more optimal result for a search query based on feedback sessions for the user.

Keywords: Data Preprocessing, Session Identification, Web log mining, Web Personalization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2023

Authors: Sufian Ashraf Mazhari, Surendra Kumar

Abstract:

Robot manipulators are highly coupled nonlinear systems, therefore real system and mathematical model of dynamics used for control system design are not same. Hence, fine-tuning of controller is always needed. For better tuning fast simulation speed is desired. Since, Matlab incorporates LAPACK to increase the speed and complexity of matrix computation, dynamics, forward and inverse kinematics of PUMA 560 is modeled on Matlab/Simulink in such a way that all operations are matrix based which give very less simulation time. This paper compares PID parameter tuning using Genetic Algorithm, Simulated Annealing, Generalized Pattern Search (GPS) and Hybrid Search techniques. Controller performances for all these methods are compared in terms of joint space ITSE and cartesian space ISE for tracking circular and butterfly trajectories. Disturbance signal is added to check robustness of controller. GAGPS hybrid search technique is showing best results for tuning PID controller parameters in terms of ITSE and robustness.

Keywords: Controller Tuning, Genetic Algorithm, Pattern Search, Robotic Controller, Simulated Annealing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3718

Authors: Guang-Ho Cha

Abstract:

This paper presents a context-sensitive media similarity search algorithm. One of the central problems regarding media search is the semantic gap between the low-level features computed automatically from media data and the human interpretation of them. This is because the notion of similarity is usually based on high-level abstraction but the low-level features do not sometimes reflect the human perception. Many media search algorithms have used the Minkowski metric to measure similarity between image pairs. However those functions cannot adequately capture the aspects of the characteristics of the human visual system as well as the nonlinear relationships in contextual information given by images in a collection. Our search algorithm tackles this problem by employing a similarity measure and a ranking strategy that reflect the nonlinearity of human perception and contextual information in a dataset. Similarity search in an image database based on this contextual information shows encouraging experimental results.

Keywords: Context-sensitive search, image search, media search, similarity ranking, similarity search.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 641

Authors: V.K.Ananthashayana, Pushpa.M.K

Abstract:

In this paper a new Joint Adaptive Block Matching Search (JABMS) algorithm is proposed to generate motion vector and search a best match macro block by classifying the motion vector movement based on prediction error. Diamond Search (DS) algorithm generates high estimation accuracy when motion vector is small and Adaptive Rood Pattern Search (ARPS) algorithm can handle large motion vector but is not very accurate. The proposed JABMS algorithm which is capable of considering both small and large motions gives improved estimation accuracy and the computational cost is reduced by 15.2 times compared with Exhaustive Search (ES) algorithm and is 1.3 times less compared with Diamond search algorithm.

Keywords: Adaptive rood pattern search, Block matching, Diamond search, Joint Adaptive search, Motion estimation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1694

Authors: Hazem M. El-Bakry, Qiangfu Zhao

Abstract:

In this paper, an approach to reduce the computation steps required by fast neural networksfor the searching process is presented. The principle ofdivide and conquer strategy is applied through imagedecomposition. Each image is divided into small in sizesub-images and then each one is tested separately usinga fast neural network. The operation of fast neuralnetworks based on applying cross correlation in thefrequency domain between the input image and theweights of the hidden neurons. Compared toconventional and fast neural networks, experimentalresults show that a speed up ratio is achieved whenapplying this technique to locate human facesautomatically in cluttered scenes. Furthermore, fasterface detection is obtained by using parallel processingtechniques to test the resulting sub-images at the sametime using the same number of fast neural networks. Incontrast to using only fast neural networks, the speed upratio is increased with the size of the input image whenusing fast neural networks and image decomposition.

Keywords: Fast Neural Networks, 2D-FFT, CrossCorrelation, Image decomposition, Parallel Processing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2179

Authors: Frederike Ohrem, Lukas Sikorski, Bastian Haarmann

Abstract:

Often the users of a semantic search application are facing the problem that they do not find appropriate terms for their search. This holds especially if the data to be searched is from a technical field in which the user does not have expertise. In order to support the user finding the results he seeks, we developed a domain-specific ontology and implemented it into a search application. The ontology serves as a knowledge base, suggesting technical terms to the user which he can add to his query. In this paper, we present the search application and the underlying ontology as well as the project EnArgus in which the application was developed.

Keywords: Information system, knowledge representation, ontology, semantic search.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1729

Authors: Syed Salman Ahmed, Zahid Halim, Rauf Baig, Shariq Bashir

Abstract:

With the rapid growth in business size, today's businesses orient towards electronic technologies. Amazon.com and e-bay.com are some of the major stakeholders in this regard. Unfortunately the enormous size and hugely unstructured data on the web, even for a single commodity, has become a cause of ambiguity for consumers. Extracting valuable information from such an everincreasing data is an extremely tedious task and is fast becoming critical towards the success of businesses. Web content mining can play a major role in solving these issues. It involves using efficient algorithmic techniques to search and retrieve the desired information from a seemingly impossible to search unstructured data on the Internet. Application of web content mining can be very encouraging in the areas of Customer Relations Modeling, billing records, logistics investigations, product cataloguing and quality management. In this paper we present a review of some very interesting, efficient yet implementable techniques from the field of web content mining and study their impact in the area specific to business user needs focusing both on the customer as well as the producer. The techniques we would be reviewing include, mining by developing a knowledge-base repository of the domain, iterative refinement of user queries for personalized search, using a graphbased approach for the development of a web-crawler and filtering information for personalized search using website captions. These techniques have been analyzed and compared on the basis of their execution time and relevance of the result they produced against a particular search.

Keywords: Data mining, web mining, search engines, knowledge discovery.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2053

Authors: Hao Wu, Jonathan L. Shapiro

Abstract:

The Bayesian Optimization Algorithm (BOA) is an algorithm based on the estimation of distributions. It uses techniques from modeling data by Bayesian networks to estimating the joint distribution of promising solutions. To obtain the structure of Bayesian network, different search algorithms can be used. The key point that BOA addresses is whether the constructed Bayesian network could generate new and useful solutions (strings), which could lead the algorithm in the right direction to solve the problem. Undoubtedly, this ability is a crucial factor of the efficiency of BOA. Varied search algorithms can be used in BOA, but their performances are different. For choosing better ones, certain suitable method to present their ability difference is needed. In this paper, a greedy search algorithm and a stochastic search algorithm are used in BOA to solve certain optimization problem. A method using Kullback-Leibler (KL) Divergence to reflect their difference is described.

Keywords: Bayesian optimization algorithm, greedy search, KL divergence, stochastic search.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1700

Authors: N. M. Nawi, M. R. Ransing, R. S. Ransing

Abstract:

The conjugate gradient optimization algorithm usually used for nonlinear least squares is presented and is combined with the modified back propagation algorithm yielding a new fast training multilayer perceptron (MLP) algorithm (CGFR/AG). The approaches presented in the paper consist of three steps: (1) Modification on standard back propagation algorithm by introducing gain variation term of the activation function, (2) Calculating the gradient descent on error with respect to the weights and gains values and (3) the determination of the new search direction by exploiting the information calculated by gradient descent in step (2) as well as the previous search direction. The proposed method improved the training efficiency of back propagation algorithm by adaptively modifying the initial search direction. Performance of the proposed method is demonstrated by comparing to the conjugate gradient algorithm from neural network toolbox for the chosen benchmark. The results show that the number of iterations required by the proposed method to converge is less than 20% of what is required by the standard conjugate gradient and neural network toolbox algorithm.

Keywords: Back-propagation, activation function, conjugategradient, search direction, gain variation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2838

Authors: Phang Chang, Phang Piau

Abstract:

Wavelet transform or wavelet analysis is a recently developed mathematical tool in applied mathematics. In numerical analysis, wavelets also serve as a Galerkin basis to solve partial differential equations. Haar transform or Haar wavelet transform has been used as a simplest and earliest example for orthonormal wavelet transform. Since its popularity in wavelet analysis, there are several definitions and various generalizations or algorithms for calculating Haar transform. Fast Haar transform, FHT, is one of the algorithms which can reduce the tedious calculation works in Haar transform. In this paper, we present a modified fast and exact algorithm for FHT, namely Modified Fast Haar Transform, MFHT. The algorithm or procedure proposed allows certain calculation in the process decomposition be ignored without affecting the results.

Keywords: Fast Haar Transform, Haar transform, Wavelet analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3140

Authors: Mario Kubek, Herwig Unger

Abstract:

Centroid terms are single words that semantically and topically characterise text documents and so may serve as their very compact representation in automatic text processing. In the present paper, centroids are used to measure the relevance of text documents with respect to a given search query. Thus, a new graphbased paradigm for searching texts in large corpora is proposed and evaluated against keyword-based methods. The first, promising experimental results demonstrate the usefulness of the centroid-based search procedure. It is shown that especially the routing of search queries in interactive and decentralised search systems can be greatly improved by applying this approach. A detailed discussion on further fields of its application completes this contribution.

Keywords: Search algorithm, centroid, query, keyword, cooccurrence, categorisation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 623

Authors: Niklas Echterhoff, Benjamin Amshoff, Jürgen Gausemeier

Abstract:

Creativity is often based on an unorthodox recombination of knowledge; in fact: 80% of all innovations use given knowledge and put it into a new combination. Cross-industry innovations follow this way of thinking and bring together problems and solution ideas from different industries. Therefore analogies and search strategies have to be developed. Taking this path, the questions where to search, what to search and how to search have to be answered. Afterwards, the gathered information can be used within a planned search process. Identified solution ideas have to be assessed and analyzed in detail for the success promising adaption planning.

Keywords: analogy building, cross-industry innovations, knowledge transfer, solution adaption.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2059

Authors: Mario Kubek, Herwig Unger

Abstract:

Due to the large amount of information in the World Wide Web (WWW, web) and the lengthy and usually linearly ordered result lists of web search engines that do not indicate semantic relationships between their entries, the search for topically similar and related documents can become a tedious task. Especially, the process of formulating queries with proper terms representing specific information needs requires much effort from the user. This problem gets even bigger when the user's knowledge on a subject and its technical terms is not sufficient enough to do so. This article presents the new and interactive search application DocAnalyser that addresses this problem by enabling users to find similar and related web documents based on automatic query formulation and state-ofthe- art search word extraction. Additionally, this tool can be used to track topics across semantically connected web documents.

Keywords: DocAnalyser, interactive web search, search word extraction, query formulation, source topic detection, topic tracking.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1648