The World of Computer Science and Information Technology Journal

Number 5, (12), 2015

8
Fingerprint Singular Point Detection via Quantization and Fingerprint Classification

Authors: Shing Chyi Chua, Eng Kiong Wong, Alan Wee Chiat Tan

Abstract

View in PDF

Number of views: 312

This paper aims to present a fingerprint singular point detection algorithm and a rule-based fingerprint classification method. The singular point detection algorithm uses a quantization approach on the orientation field of the fingerprint image and seeks to locate the core and delta points via the changes of the gray levels around a 2x2 window. It has been found that with the application of an edge-trace-cum-core-delta-pairing algorithm and a merging-and-pruning heuristic as the post-processing steps, spurious singular points are removed and the final singular points are then used for classification. Fingerprint classification on NIST-4 database by rule-based method utilizes the number of singular points and three key geometry features to perform 5-class as well as 4-class classification using success rate (the accuracy) as the performance measure. It has been found to achieve 86.5% and 92.15% of success rate, respectively. The study has thus find the application of the new singular point detection algorithm via quantization and the rule-based classification to be promising as many of the fingerprint images in the NIST-4 database have been reported as poor quality, i.e. 22.35%. Keywords-singular point; fingerprint singular point detection; fingerprint classification; quantization; fingerprint

Number 5, (11), 2015

7
An Empirical Analysis of Multiple Level Association Rules Mining Method for Feature Extraction

Authors: Ruchika Yadav, Kanwal Garg

Abstract

View in PDF

Number of views: 328

Mining multiple level association rules in large databases is core area of data mining. Discovering these rules is favorable to the accurate and appropriate decision made by decision makers. Frequent patterns discovery is the key process in multiple levels association rule mining. One of the challenges in developing multiple level association rules mining approaches is to implement an iterative procedure to find association rule, which takes a intricate transaction process. Moreover, the offered mining methods cannot perform proficiently due to repetitive disk access overhead. Due to this, a novel method named MLTransTrie is presented in this paper. It can efficiently discover the association rules at multiple levels of abstraction in large databases and provide the more exact and precise information. The focus is on the comparative exploration and performance assessment of the MLTransTrie algorithm that generates multiple trie structure for all levels in one database scan. For this, the performance of this new algorithm MLTransTrie is analyzed and compared with LWFT and MLT2_L1 algorithm on diverse class of datasets (four real world dataset) and parameters. Keywords- Concept Hierarchy; Confidence; Frequent Patterns; Minimum Support; MLTransTrie Implementation; Multiple-Level Association Rule.

Number 5, (10), 2015

10
Data Quality Based Data Integration Approach

Authors: Mohamed Samir Abdel-Moneim, Ali Hamed El-Bastawissy, Mohamed Hamed Kholief

Abstract

View in PDF

Number of views: 332

Data integration systems (DIS) are systems where query answers are collected from a set of heterogeneous and autonomous data sources. Data integration systems can improve results by detecting the quality of the data sources and retrieve answers from the significant ones only. The quality measures of the data in the data sources not only help in determining the significant data sources for a given query but also help data integration systems produce results in a reasonable amount of time and with less errors. In this paper, we perform an experiment that shows a mechanism used to calculate and store a set of quality measures on data sources. The quality measures are, then, interactively used in selecting the most significant candidates of data sources to answer users’ queries. The justification and evaluations are done using amalgam and THALIA benchmarks. We show that our approach dramatically improves query’s answers. Keywords-component data integration; quality measures; data sources; query answers; user preferences.

Number 4,(9), 2014

6
Discrete Wavelet Transform Based Algorithm for Recognition of QRS Complexes

Authors: Rachid HADDADI, Elhassane ABDELMOUNIM, Mustapha EL HANINE, Abdelaziz BELAGUID

Abstract

View in PDF

Number of views: 560

this paper proposes the application of Discrete Wavelet Transform (DWT) to detect the QRS (ECG is characterized by a recurrent wave sequence of P, QRS and T-wave) of an electrocardiogram (ECG) signal. Wavelet Transform provides localization in both time and frequency. In preprocessing stage, DWT is used to remove the baseline wander in the ECG signal. The performance of the algorithm of QRS detection is evaluated against the standard MIT BIH (Massachusetts Institute of Technology, Beth Israel Hospital) Arrhythmia database. The average QRS complexes detection rate of 98.1 % is achieved.

9
Analysis of Performance Parameters in Wireless Networks by using Game Theory for the non Cooperative Slotted Aloha Enhanced by ZigZag Decoding Mechanism

Authors: Abdellah ZAALOUL and Abdelkrim HAQIQ

Abstract

View in PDF

Number of views: 328

In wireless communication networks, when the workload increases, sources become more aggressive at the equilibrium in the game setting in comparison with the team problem by using slotted Aloha mechanism. Consequently, more packets are in collision and are lost. To reduce these phenomena and to enhance the performance of the networks, we propose to combine ZigZag decoding approach with non cooperative slotted Aloha mechanism. This approach was introduced in our previous work based on the cooperative slotted Aloha mechanism. The obtained results showed that this approach has significantly improved the cooperative slotted Aloha mechanism and gave best results for the throughput and delay. In this paper, we analyze the impact of combining non cooperative slotted Aloha and ZigZag Decoding. We model the system by a two dimensional Markov chain that integrates the effect of ZigZag decoding. The states of the Markov chain describe the number of backlogged packets among the users. We use a stochastic game to achieve our objective; we evaluate and compare the performances parameters of the proposed approach with those of a simple slotted Aloha mechanism. All found results show that our approach improves the performance parameters of the system. Keywords- Non Cooperative Slotted Aloha; Markov Process; MAC Protocol; ZigZag Decoding; Performance Metrics.

Number 4,(8), 2014

5
Probabilistic PCA Mixture under Variance Preservation

Authors: Mohamed Nour I. Ismail, Mohamed El-Hafiz Mustafa Muse

Abstract

View in PDF

Number of views: 420

modeling data heterogeneity by a mixture of local models and exploiting the correlation in the localized data subsets to reduce their subspace dimensionalities has been realized in many mixture models; like PCA mixture and FA mixture models. Determining the number of local models as well as the proper dimensionality for each subspace (local model space) are the most difficult questions of these models. Instead of using fixed ad-hoc dimensionality for all local models, this paper proposes using a global preserved variance percentage value to estimate the dimensionality that retains the given variability percentage in each subspace. We test the proposed method on classifying handwritten digit by a mixture of Probabilistic PCA model, the result shows that the proposed method outperforms fixed dimensionality probabilistic PCA mixture model.

8
Adaptive Activation Function for Isolated Digit Recognition Based on Speaker Dependent System

Authors: Ummu Salmah, M.H, Siti Mariyam, S., Saira Banu, O.K., Nor Azah,

Abstract

View in PDF

Number of views: 352

An automatic speech recognition (ASR) system has been the goal in speech research for more than 6 decades. This study focuses on developing the robustness of the MLP neural network for the Malay isolated digit recognition system by proposing a simple novel approach. An adaptive sigmoid function is implemented to achieve this objective. A typical or fixed sigmoid function method is used in the learning phase. In the recognition phase, an adaptive sigmoid function is employed. In this sense, the slope of the activation function is adjusted to gain highest recognition rate. The outcome of the simulation reveals that adaptive sigmoidal function offers a number of advantages over traditional fixed sigmoid function, resulting in better generalization performance. The proposed approach implicates ASR is applicable for the task on Malay language continuous speech and the speaker independent task to fulfill the ultimate goal in speech technology, towards natural ASR.

4 5 6 7 8 9 10