eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
1
18
10.21608/ijicis.2016.19821
19821
Original Article
A PROPOSED METHOD FOR INCREASE ACCURACY OF CLASSIFICATION P300 SPELLER
A. Khalaf
1
M. El-Desouky
2
M. Rashad
3
Faculty of Computers and Information, Mansoura University, Egypt
Faculty of Computers and Information, Mansoura University, Egypt
Faculty of Computers and Information, Mansoura University, Egypt
A P300 speller is one of applications the brain computer interface (BCI), introduced by Farwell and Donchin in 1988 N. In this paper proposed a new method for increase the accuracy of classification P300 speller. Uses a dataset for (16) healthy subjects. The new method including is feature extraction using Principle Component Analysis (PCA), and the classification using Support Vector Machine Linear (SVML). As it was calculated performance and activity of each electrode whether correlated or uncorrelated of speller task. Show that the proposed method is accurate andefficient.
https://ijicis.journals.ekb.eg/article_19821_40f54fd76aa571fb83e5fa9bbf5ddb91.pdf
Brain Computer Interface (BCI)
P300 speller
Accuracy of classification P300
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
19
28
10.21608/ijicis.2016.19822
19822
Original Article
CLASSIFICATION OF LOW QUALITY IMAGES USING CONVOLUTIONAL NEURAL NETWORK AND DEEP BELIEF NETWORK
E. El-Ashmony
1
M. El-Dosuky
2
Samir Elmougy
3
Department of Computer Science, Faculty of Computers and Information,Mansoura University, Mansoura 35516, Egypt
Department of Computer Science, Faculty of Computers and Information,Mansoura University, Mansoura 35516, Egypt
Department of Computer Science, Faculty of Computers and Information,Mansoura University, Mansoura 35516, Egypt
Low quality images become more challenge and core problem in recent decade because of the ambiguity of contents of them. Convolutional deep neural networks are used for solving this problem. In this work, we used a combination of convolutional neural network and deep belief network to construct an efficient model able to classify low quality images. This model has the capability in extracting effective features from low quality images. Data augmentation is used through this model to increase the accuracy of the system. Scikit-Learn python library is used in implementation the system on STL-10 dataset. The results showed that the proposed model increase the accuracy of the system by 0.20%.
https://ijicis.journals.ekb.eg/article_19822_24d5a5614030ffe40b5d10669cce52dc.pdf
Convolutional deep neural networks
Deep belief network
Low quality images
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
29
40
10.21608/ijicis.2016.19824
19824
Original Article
REDUCING ATTRIBUTES of FACEBOOK USERS USING ROUGH SET THEORY
W. Abdallah
1
S. Sarhan
2
Samir Elmougy
3
Dept. of Computer Science, Faculty of Computers and Information, Mansoura University, Mansoura 35516, Egypt.
Dept. of Computer Science, Faculty of Computers and Information, Mansoura University, Mansoura 35516, Egypt.
Dept. of Computer Science, Faculty of Computers and Information, Mansoura University, Mansoura 35516, Egypt
Using social networks have become one of the daily activities that billions of peoples around the world do. So, great research efforts had been done to analyze and understand these virtual communities. Among other things, link prediction is a paramount task to analyze and understand these social networks. In this paper, we investigate link prediction problem using rough set theory to discard the irrelevant attributes that could be found in the profiles of Facebook users and the proposed workinduces accuracy 97.79%.
https://ijicis.journals.ekb.eg/article_19824_96194586a7706237fc03256d38d5e6e6.pdf
Link Prediction
Social networks
Rough set theory
facebook
Self-Organization Map
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
41
53
10.21608/ijicis.2016.19823
19823
Original Article
REDUCING ERROR RATE OF DEEP LEARNING USING AUTO ENCODER AND GENETIC ALGORITHMS
F. Habeeb
1
Sherihan Abuelenin
2
Samir Elmougy
3
Faculty of Computers and Information, Mansoura University, Egypt
Faculty of Computers and Information, Mansoura University, Egypt
Faculty of Computers and Information, Mansoura University, Egypt
Deep Learning (DL) techniques are considered as one of machine learning classes that model hierarchical abstractions in data input with the assistance of multiple layers. DL techniques have accomplished high performance in computer vision, natural language processing and automatic speech recognition. DL combines lower modules for classifier output and raw features input to produce new features at hierarchy higher layer. Deep Auto Encoder (DAE) is a DL aims to represent data to be utilized for rebuilding and classification. It is considered as one of the powerful algorithms in DL that gives higher accuracy and best performance. The proposed method in this work is based on using DAE and Genetic Algorithm (GA) through applying split-training and merging algorithms for DL. First, the network is divided into two initialized networks using DAE. Second, both of these networks were merged using GA. This proposed approach was evaluated based on the Mixed National Institute of Standards and Technology (MNIST) dataset and the obtained results showed that it achieve a higher performance and lower error rate in the classification.
https://ijicis.journals.ekb.eg/article_19823_68459f00c75d816b0e0a3ec0fa6b533b.pdf
Deep Auto Encoder
Genetic Algorithm
Machine Learning
Deep learning
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
55
63
10.21608/ijicis.2016.19825
19825
Original Article
FACIAL EXPRESSION RECOGNITION BASED ON PRINCIPAL COMPONENTS ANALYSIS
A. Hewa
1
O. Nomir
2
A. Saleh
3
Department of Computer Science, Faculty of Computer and Information, Mansoura University ,Egypt
Department of Computer Science, Faculty of Computer and Information, Mansoura University ,Egypt
Department of Computer Science, Faculty of Computer and Information, Mansoura University ,Egypt
Recognizing facial expression is one of the most effective applications of image processing and has obtained great attention in latest years. A recognition system for facial expression is a computer based application which detects an individual facial expression for the purposes of authentication, criminal identification, passport verification, estimating age, and various other purposes. In this study, we propose a human recognition system based on facial expression. The system depends on extracting features using Principal Component Analysis (PCA) which later used in the training and recognition steps. The system is able to recognize diverse facial expressions such as Neutral, Anger, Disgust ,Fear, Happy, Sad and Surprise. The primary objective of this study is to improve the efficiency and to achieve better recognition rate using Support Vector Machine (SVM). The system is evaluated using the registered JAFFE Dataset of face images. The results show that or proposed system is robust and maintain high recognition rate.
https://ijicis.journals.ekb.eg/article_19825_f8ab5be8392f4a8ef0ac35c3af3f3846.pdf
Facial Expression
PCA
Feature Extraction
SVM
recognition rate
JAFFE Face Dataset
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
65
78
10.21608/ijicis.2016.19826
19826
Original Article
FAULT NODE RECOVERY ALGORITHM FOR ENHANCINGTHE LIFETIME OF AWIRELESS SENSOR NETWORK
S. Dawood
1
S. Abuelenin
2
A. Atwan
3
Faculty of computers and Information, Mansoura University, Egypt
Faculty of computers and Information, Mansoura University, Egypt
Faculty of computers and Information, Mansoura University, Egypt
Wireless Sensor Network (WSN) is type of network which consists of collection of tiny device called sensors nodes. In real wireless sensor networks, the sensor nodes use battery power supplies and thus have limited energy resources. This paper decrease the number of fault node and loss data and number of routing in the network based on enhanced Grade Diffusion with using Shortest Best Path .In the simulation, number of hop, power consumption, fault detection accuracy and time number of neighbour nodes measure the proposed algorithm. The proposed algorithm is also compared with distributed fault detection (DFD) and fault node recovery (FNR).
https://ijicis.journals.ekb.eg/article_19826_a6d9cc4d083970a7fead5e8b71e0c9f4.pdf
WSNs
failure detection
failure recovery algorithm
Grade Diffusion
and dead node Shortest Save Path
and Fault Node Recovery
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
79
87
10.21608/ijicis.2016.19829
19829
Original Article
A SYSTEM FOR ACUTE LEUKEMIA CELLS SEGMENTATION AND CLASSIFICATION
R. Mohammed
1
O. Nomir
2
I. I. Khalifa
3
T. Hamza
4
Computer Science Department, faculty of computer and informatics Mansoura University , Egypt
Computer Science Department, faculty of computer and informatics Mansoura University , Egypt
Computer Science Department, faculty of computer and informatics
Computer Science Department, faculty of computer and informatics
This research paper presents a system for the acute leukemia blast cells segmentation and classification. The research objective is to generate the features characterizing normal and infected cells. The proposed system consists of one segmentation method and one classification method of acute leukemia. The features extracted from the cell and adopted features are used as the input signals to the Multi Layer Perception (MLP) neural network classifier. The experimental results show that our proposed system is robust and effective in identifying acute leukemia blast cells.
https://ijicis.journals.ekb.eg/article_19829_003cce288901d1ac104caef418b13cd0.pdf
mage Segmentation
ALL
RGB
C-Y color model
features extraction
MLP
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
89
97
10.21608/ijicis.2016.19830
19830
Original Article
XML ABSTRACTIVE SUMMARY APPROACH
H. Elmadany
1
M. Alfonse
2
M. Aref
3
Computer Science Department, Faculty of Computer and Information Sciences, Ain Shams University, Cairo, Egyp
Computer Science Department, Faculty of Computer and Information Sciences, Ain Shams University, Cairo, Egyp
Computer Science Department, Faculty of Computer and Information Sciences, Ain Shams University, Cairo, Egyp
Text summarization saves both time and effort required to manage a vast amount of information. The need to summarize text is increased. This paper introduces a XML Abstractive Summary (XAS) approach to summarize text in the format of XML document that is called XML summarization. XAS approach is considered a new attempt to produce abstractive summary for the xml document regarding to performance, size and accuracy. The output document is a concise and readableversion for the original one.
https://ijicis.journals.ekb.eg/article_19830_bb78b32069822aab1dddb5d3f18590ae.pdf
XML Summarization
Abstractive Summarization
Ranking
Rich Semantic Graph
Functional Dependancy
eng
Ain Shams University, Faculty of Computer and Information Science
International Journal of Intelligent Computing and Information Sciences
1687-109X
2535-1710
2016-10-01
16
4
99
108
10.21608/ijicis.2016.30054
30054
Original Article
A PROPOSED FRAMEWORK FOR THE INTEGRATION OF E-GOVERNMENT DATA AND SERVICES
Amira Rezk
amira_rezk@mans.edu.eg
1
Sherif Barakat
sheiib@mans.edu.eg
2
Sirwan Abdullah
mamo.saber@yahoo.com
3
Department of Information System, Faculty of Computer and Information Sciences, Mansoura University, Egypt
Department of Information System, Faculty of Computer and Information Sciences, Mansoura University, Egypt
Department of Information System, Faculty of Computer and Information Sciences, Mansoura University, Egypt
https://ijicis.journals.ekb.eg/article_30054_c6e0a30c721f1e489aeea0398111614d.pdf