The Analysis of Traffic of IP Packets using CGH. Self Organizing Map

Size: px
Start display at page:

Download "The Analysis of Traffic of IP Packets using CGH. Self Organizing Map"

Transcription

1 2015 International Conference on Computational Science and Computational Intelligence The Analysis of Traffic of IP Packets using CGH Self Organizing Maps Hiroshi Dozono Department of Advanced Fusion Saga University 1-Honjyo Saga JAPAN Nozomu Okada FUJITSU GENERAL JAPAN Abstract Recently, many packet analysis methods are proposed for the threat of the increasing malware. In this research, the method for the analysis of the IP packet based on SOM(Self- Organizing Map) using the frequency of appearance of IPpackets as input vectors is proposed. The performance of this method is examined by the experiment of detecting the packets generated by emulating malware behaviors. I. INTRODUCTION As the growth of communication networks, the security issues concerning network are increasing. As a attacking technique to the server, DOS/DDOS attacks are often used. Under these attacks, a large number of packets are sent to the server to increase the load of the server, and to disturb the normal operation. To defend the server from such attacks, the analysis of the packet traffic is effective. We proposed a method for analyzing the packet traffics using Self Organizing Map(SOM)s[1]. The packet data can be captured using WinPcap, which is the packet driver for WINDOWS OS. We proposed a visualization method of packet traffics using SOM and conducted some experiments for detecting the attacks. As the previous work using SOM for network traffic analysis, in [2] the large amount of logs from IDS systems were analyzed and the information of each packet was learned using SOM. In [3], the occurrence rate of the IP address in the list of IP address was analyzed by using SOM. With expanding the scope to machine learning, many machine learning method including neural networks are applied to the detection of abnormal traffics originated from malware using the KDD CUP 99 data as sample set[4]. In this paper, we propose a method for analyzing packet traffics using Computer Generated Hologram(CGH) SOM[5] as a method for analyzing big data. In the previous study mentioned before, we propose a method using conventional SOM and Pareto learning SOM. On the other hand, we proposed CGH SOM for clustering 2 dimensional images and 3 dimensional objects. In this paper, CGH SOM is applied to the analysis of packet traffic to extract the feature of the packet traffics effectively. Additionally, we conducted the experiments of detecting unauthorized traffic using unsupervised learning considering the situation of unknown attacks. II. SELF ORGANIZING MAP(SOM) Self-Organizing Map (SOM)[2] is the model of the neurologic function of the cerebral cortex developed by T.Kohonen. As the class of neural networks, SOM is the feedforward type of two hierarchical without the interlayer using algorithms of unsupervised learning. SOM converts a nonlinear and statistics relations that exist between higher dimension data into the image with a simple geometrical relation. They can usually be used to make a higher dimension space visible because it is displayed as the lattice of the neuron of two dimensions. Moreover, it becomes easy to be able to visualize higher dimension information because it is possible to cluster without the preliminary knowledge, and to understand the relations among these multidimensional data intuitively for human. SOM is basically a Fig. 1. Self Organizing Map network in the double-layered structure in the input layer and the competitive layer as shown in Fig.1. However, there are not connections between neurons in the same layer. The first layer is input layer x of n dimension, and the second layer is called a competitive layer, and is generally two dimension array for visualizing output data. It is assumed that it has the neuron that input data is given by n-dimensional real number vector x =(x 1,x 2,..., x n ) and two dimensions SOM were arranged on the grid point of m m(= M)piece. Input data x is presented to all neurons. The neuron located in (j, i)in two dimension lattice array has w j,i = (w 1 j,i,w2 j,i,..., wn j,i ) /15 $ IEEE DOI /CSCI

2 weight vectors which are tunable corresponding to the input vectors. This w j,i is called as reference vector. The algorithm can be divided into two phases in a competitive phase and the cooperative phase. First of all, the neuron of the closest,in a word, the winner neuron, is selected at a competitive phase. Next, the winner s weight are adjusted at the cooperation phase as well as the lattice near the winner. TABLE II SETTING OF THE EXPERIMENT Map size Number of packets Group size 500 Overlap of group 200 Protocol TCP III. PACKET ANALYSIS USING SOM A. Packet capture In this experiments, the packet data which is captured in our laboratory is used. Packet data is captured by the program which uses WinPcap library which is build for WINDOWS OS. The packets are captured in promiscuous mode. However, the unauthorized packets are already captured by 3 firewalls which is located between the laboratory and external internet. Thus we can get only authorized packets. B. Packet analysis using frequencies It needs large computational costs to analyze each packet independently using SOM because the number of packets is very large. And Dos and DDos attacks are collective behavior of the packets. From these reasons, the statistical information of the group of the packets is used. As the statistical information, the frequency of the code in the group of packets is used. Frequencies of the specific data are often used to visualize the big data by SOM[6]. Table I shows the contents of the input vector which is composed of the frequencies of the elements in the packet. For each 8[bit](1 byte) of 32[bit] IP address, TABLE I CONTENTS OF INPUT VECTOR Packet element dimension Source IP adress Destination IP address Source port number 20 Destination port number 20 Packet length 7 Payload 256 the frequency of each code is counted. For example, if the source IP address is , 133th of 1st 256 elements, 49th of 2nd 256 elements, 30th of 3rd 256 elements and 1st of 4th 256 elements are counted up respectively. As for port number, the frequencies of 20 port numbers which are frequently used by malware are counted. Packet length is classified to 7 classes, and the frequency of each class is counted. For payload data, the frequency of each code is counted. C. Mapping the traffic of authorized packets At first, the experimental result of mapping the traffic of the captured packets which were filtered by firewalls. The setting of the experiment is shown in Table 2. Fig.2 shows the map organized by SOM. Each dot represents a group of packets, and the shade of the color denotes the sequential order of the group. The color is continuously changing on the map, and it Fig. 2. Mapping of the authorized packet means that authorized traffic of the packets are continuously arranged on the map. D. Experiments of detection of the traffic including unauthorized packets Next, Experiments of detection of the traffic including unauthorized packets is conducted. As mentioned before, the unauthorized packets cannot reach to our laboratory, the pseudo unauthorized packets which simulate Dos and DDos attacks are used in these experiments. Dos packets and DDos packets are generated with setting the source IP addresses as fixed number and random number respectively. After initial learning using authorized packets, the groups of packets including unauthorized packets are given as input of SOM. The groups whose mapped distance from previous group of authorized packets are larger than 2σ are detected as the group including unauthorized packet, where σ is standard deviation of the distance before the group is processed. During the attacks, 80% of the packets are unauthorized packets, and 3 experiments are conducted for each case. For each cases, 5 experiments are conducted, and average of the experiments is presented. Table 3 and Table 4 shows the results for simulated Dos and DDos attacks with changing overlap size of the group respectively. The results vary with the overlap size, TABLE III EXPERIMENTAL RESULTS FOR SIMULATED DOS ATTACKS Overlap size Specificity Sensitivity and specificity is too low for almost cases. To improve the accuracy, CGH SOM is applied

3 TABLE IV EXPERIMENTAL RESULTS FOR SIMULATED DDOS ATTACKS Overlap size Specificity Sensitivity IV. COMPUTER GENERATED HOLOGRAM (CGH) SOM CGH-SOM is the composition of SOM and matched filtering using CGH, and can be applied to cluster the images data and 3D objects. Especially, matched filtering is powerful for the matching of 2 dimensional images including shifts. 1) CGH: Hologram is the record of the interference patterns of reflected beam of the object and reference beam. Assuming that the reflected beam of the object and reference beam are O and R respectively, and that the transmittance of the dry plate is in proportion to the amplitude of the exposure beam, the distribution of the transmittance is given as follows. O 2 + R 2 (1) CGH processes the optical operation of hologram in computer. The hologram is calculated based on the spatial information of the object as shown in Fig.3. In Fig.3, Taking the amplitude of Fig. 3. Calculation of CGH the reference beam R and object beam O as A 1 and A 2, and the cosine of the axis of the coordinate (xyz) as cosα x1,cosα y1 and cosα z1, R and O at (x, y, z) =(x 0,y 0, 0) is given as follows. R(x 0,y 0, 0) = A 1 exp( ik(x 0 cosα x1 + y 0 cosα y1 ))(2) O(x 0,y 0, 0) = A 1 exp( ikz 2 ik ((x 0 x 2 ) 2 z 2 2z 2 (3) +(y 0 y 2 ) 2 )) (4) where k is wave number. From these equations, the interference pattern I is given as follows. I(x 0,y 0, 0) = R + O 2 (5) = A A2 2 z 2 2 This equation represents CGH. (6) +2 A 1A 2 z 2 (7) cos(k(d (x 0 cosα x1 + y 0 cosα y1 )))(8) 2) Matched filtering: Matched filtering is a method for pattern recognition using hologram, and the correlation between 2 images can be calculated as numerical value. For images, the Fourier transform hologram is generally used as the filter. Assume that the object beam used to generate hologram and the object beam of the input image for matching are O 1 and O 2 respectively. As the distribution of the transmittance is R + O 1 2, the distribution of the output beam with projecting the object beam O 2 is given as follows. R + O 1 2 O 2 (9) =( R 2 + O 1 2 ) O 2 + R O 1 O 2 + R O1 O 2 (10) where denotes complex conjugate. The 3rd term of this equation that represent reference beam component becomes O 1 2 R, if the image that is used to generate hologram and input image is same. Because the reference component is plane wave, the peak pattern called as correlation spot is detected with transmitting through lens, and the amplitude of the peak represents the similarity between the images. All of the optical process is simulated in computer in CGH, and the lens is simulated by Fourier transform in CGH. 3) The algorithm of CGH-SOM: CGH-SOM is the self organizing map which is composed of the units using CGH as memories. The learning algorithm is almost the same as that of conventional SOM with batch updates. CGH-SOM Algorithm Step1 : Preprocessing of images For each image Data D i, calculate the Fourier transform hologram F i. Step 2 : Initialization of the map For each unit U ij on the map, generate the initial hologram h ij randomly. Step 3: Matching Match the holograms on the map and Fourier transform hologram of image data. For each h ij, for each F k, generate the matching image F ( h ij F k ), and calculate P max ij k which is the maximum amplitude of the correlation spot respectively. Step 4: Finding winner For each image D k, find the winner unit W ij k using the matching results P max ij k. Step 5: Updating the units For each winner W ij k, update the hologram h ij and its neighbors using the following equation. h ij = h ij + ηfn(d)(f k hij) (11) where η is learning coefficient and fn(d) is the neighboring function. Repeat Step 3 - Step 5 by decreasing the size of neighbors and learning coefficient in pre-defined iterations In step 1, the Fourier transform of the images are computed before applied to SOM because the computational costs are very large. In step 3, the input object data is matched with the holograms on the map based on the matching method of

4 hologram processing. In step 4 and step 5,standard equation of SOM can be applied for updating holograms on the map, because the holograms can be superposed by simply summing the holograms. Fig.4 shows an example of the map organized by CGH- SOM. The bitmap patterns of the character BCDEFIOQT are used as input data. As shown in Fig.4, similar patterns are arranged closely on the map. Fig. 5. Input image converted from a packets A. Mapping the traffic of authorized packets Fig.5 shows the experimental result of mapping the groups of authorized packets using CGH-SOM. As same as the Fig. 4. Example of CGH SOM V. PACKET ANALYSIS USING CGH SOM In section 3, the experimental results using SOM are mentioned. However, the accuracy is not significant. SOM uses simple Euclidian distance for matching. CGH SOM uses matched filtering of CGH for matching, and it is sensitive to abnormal data. Additionally, 2-stage SOM is applied to payload data. Payload data is clustered in the 1st stage SOM, and the resulting map of 1st stage is clustered in 2nd stage CGH-SOM. Compared with using the conventional vectorial SOM in 2nd stage, CGH-SOM is considered to be able to extract the features which are organized by 1st stage SOM because it can process 2 dimensional data. 4) Composition of input image: To apply CGH-SOM, the input data is needed to be converted to image data. in the previous experiments, the group of packets is converted to a input vector. On the other hand, each packet is converted to the input data of CGH-SOM in the following experiments. Fig.5 shows the input image which is converted from a packet. The image size is 16 16, and each pixel is 256 grayscale. The 8 bytes binary values of sender and destination IP addresses are stored in 6 pixels of column. The size of the column represents the weight of the data. Port number and header information which is comprised of header length, TTL and packet length are given 2 6 pixels and 4 6 pixels respectively. The payload data is the frequencies of 6[byte] codes in the payload of each packets which were organized in units by 1st stage SOM. 1st stage SOM is organized by using the payload data of normal traffic packets beforehand. Fig. 6. Mapping of the authorized packet using CGH-SOM mapping result of conventional SOM using groups of data as input vectors, the packets are arranged continuously on the map. B. Experiments of detection of the traffic including unauthorized packets Next, experiments of detection of the traffic including unauthorized packets are conducted. The method of this experiment is almost the same as those using conventional SOM except CGH-SOM uses each packet as input data. Table 5 and Table 6 show the results for Dos and DDos attacks respectively. The sizes of group are taken 150 and 50 for Dos and TABLE V EXPERIMENTAL RESULTS FOR SIMULATED DOS ATTACKS Rate of the unauthorized packets Specificity Sensitivity DDos attacks respectively, which show the best results in

5 TABLE VI EXPERIMENTAL RESULTS FOR SIMULATED DDOS ATTACKS Rate of the unauthorized packets Specificity Sensitivity the previous experiments. As for Dos attacks, the accuracy is improved for the same rate(0.8) of unauthorized packets and even for the smaller rates of unauthorized packets. As for DDos attacks, the accuracy becomes worse for all cases. VI. CONCLUSION In this paper, the methods for analyzing packet traffic using SOM and CGH-SOM are proposed. Using the frequencies of codes in the header and payload ss the statistical information of the group of packets and the input data converted from each packet, the traffic of packets is visualized as the continuous flow on the map by SOM and CGH-SOM respectively. The experiments of detecting pseudo Dos and DDos attacks are conducted, and CGH-SOM can improve the accuracy for Dos attack, however conventional SOM is still superior to detect DDos attacks. As for further works, the comparison with other method should be conducted, and the accuracy of detection should be improved much more to use this method practically. ACKNOWLEDGMENT This work was supported by JSPS KAKENHI Grant Number REFERENCES [1] H. Dozono and T. Kabashima,et.al. Analysis of Packet Traffics and Detection of Abnormal Traffics Using Pareto Learning Self Organizing Maps, ICONIP 2010, Part II, LNCS 6444, pp , Springer, [2] K. Ohkouchi and K. Rikitake,et.al. A Study on Network Incident Analysis Using Self Organizing Maps, Proceedings of the 2006 Symposium on Cryptography and Information Security, [3] K. Kanenishi and K. Togawa,et.al. Aberrant Detection from Behavior of Campus Network Traffic, Journal of Academic Computing and Networking (13), 74?83, 2009 [4] Mahbod Tavallaee, et.al,. A Detailed Analysis of the KDD CUP 99 Data Set, Proceedings of the 2009 IEEE Symposium on Computational Intelligence in Security and Defense Applications, 2009 [5] H. Dozono and A. Tanaka,et.al. Mapping of the 3D Objects Using Computer Generated Hologram SOM, Advances in Self-Organizing Maps (WSOM2011), , Springer, 2011 [6] T. Abe and T. Ikemura,et.al. A Novel Bioinfor- matics Strategy for Phylogenetic Study of Genomic sequence Fragments: Self Organizing Map (SOM) of Oligonu- cleotide Frequencies, Proceedings of 5th Workshop on Self Organizing Maps, ,

Visualization of the Packet Flows using Self Organizing Maps

Visualization of the Packet Flows using Self Organizing Maps Visualization of the Packet Flows using Self Organizing Maps HIROSHI DOZONO Saga University Faculty of Science and Engineering 1 Honjyo Saga Saga JAPAN hiro@dna.ec.saga-u.ac.jp TAKERU KABASHIMA Saga University

More information

Two-step Modified SOM for Parallel Calculation

Two-step Modified SOM for Parallel Calculation Two-step Modified SOM for Parallel Calculation Two-step Modified SOM for Parallel Calculation Petr Gajdoš and Pavel Moravec Petr Gajdoš and Pavel Moravec Department of Computer Science, FEECS, VŠB Technical

More information

Robustness of Selective Desensitization Perceptron Against Irrelevant and Partially Relevant Features in Pattern Classification

Robustness of Selective Desensitization Perceptron Against Irrelevant and Partially Relevant Features in Pattern Classification Robustness of Selective Desensitization Perceptron Against Irrelevant and Partially Relevant Features in Pattern Classification Tomohiro Tanno, Kazumasa Horie, Jun Izawa, and Masahiko Morita University

More information

Slide07 Haykin Chapter 9: Self-Organizing Maps

Slide07 Haykin Chapter 9: Self-Organizing Maps Slide07 Haykin Chapter 9: Self-Organizing Maps CPSC 636-600 Instructor: Yoonsuck Choe Spring 2012 Introduction Self-organizing maps (SOM) is based on competitive learning, where output neurons compete

More information

Function approximation using RBF network. 10 basis functions and 25 data points.

Function approximation using RBF network. 10 basis functions and 25 data points. 1 Function approximation using RBF network F (x j ) = m 1 w i ϕ( x j t i ) i=1 j = 1... N, m 1 = 10, N = 25 10 basis functions and 25 data points. Basis function centers are plotted with circles and data

More information

Invited Paper. Nukui-Kitamachi, Koganei, Tokyo, , Japan ABSTRACT 1. INTRODUCTION

Invited Paper. Nukui-Kitamachi, Koganei, Tokyo, , Japan ABSTRACT 1. INTRODUCTION Invited Paper Wavefront printing technique with overlapping approach toward high definition holographic image reconstruction K. Wakunami* a, R. Oi a, T. Senoh a, H. Sasaki a, Y. Ichihashi a, K. Yamamoto

More information

Effect of Grouping in Vector Recognition System Based on SOM

Effect of Grouping in Vector Recognition System Based on SOM Effect of Grouping in Vector Recognition System Based on SOM Masayoshi Ohta Graduate School of Science and Engineering Kansai University Osaka, Japan Email: k287636@kansai-u.ac.jp Yuto Kurosaki Department

More information

Figure (5) Kohonen Self-Organized Map

Figure (5) Kohonen Self-Organized Map 2- KOHONEN SELF-ORGANIZING MAPS (SOM) - The self-organizing neural networks assume a topological structure among the cluster units. - There are m cluster units, arranged in a one- or two-dimensional array;

More information

Securing of Two and Three Dimensional Information Based on In-line Digital Holography

Securing of Two and Three Dimensional Information Based on In-line Digital Holography Securing of Two and Three Dimensional Information Based on In-line Digital Holography Hesham Eldeeb Computer & System Department Electronic Research Institute National Research Center, Dokki, Giza Cairo,

More information

Unsupervised Clustering of Web Sessions to Detect Malicious and Non-malicious Website Users

Unsupervised Clustering of Web Sessions to Detect Malicious and Non-malicious Website Users Unsupervised Clustering of Web Sessions to Detect Malicious and Non-malicious Website Users ANT 2011 Dusan Stevanovic York University, Toronto, Canada September 19 th, 2011 Outline Denial-of-Service and

More information

Influence of Neighbor Size for Initial Node Exchange of SOM Learning

Influence of Neighbor Size for Initial Node Exchange of SOM Learning FR-E3-3 SCIS&ISIS2006 @ Tokyo, Japan (September 20-24, 2006) Influence of Neighbor Size for Initial Node Exchange of SOM Learning MIYOSHI Tsutomu Department of Information and Knowledge Engineering, Tottori

More information

Simulation of WSN in NetSim Clustering using Self-Organizing Map Neural Network

Simulation of WSN in NetSim Clustering using Self-Organizing Map Neural Network Simulation of WSN in NetSim Clustering using Self-Organizing Map Neural Network Software Recommended: NetSim Standard v11.1 (32/64bit), Visual Studio 2015/2017, MATLAB (32/64 bit) Project Download Link:

More information

MEFT / Quantum Optics and Lasers. Suggested problems from Fundamentals of Photonics Set 1 Gonçalo Figueira

MEFT / Quantum Optics and Lasers. Suggested problems from Fundamentals of Photonics Set 1 Gonçalo Figueira MEFT / Quantum Optics and Lasers Suggested problems from Fundamentals of Photonics Set Gonçalo Figueira. Ray Optics.-3) Aberration-Free Imaging Surface Determine the equation of a convex aspherical nonspherical)

More information

Modular network SOM : Theory, algorithm and applications

Modular network SOM : Theory, algorithm and applications Modular network SOM : Theory, algorithm and applications Kazuhiro Tokunaga and Tetsuo Furukawa Kyushu Institute of Technology, Kitakyushu 88-96, Japan {tokunaga, furukawa}@brain.kyutech.ac.jp Abstract.

More information

Self-Organizing Maps for cyclic and unbounded graphs

Self-Organizing Maps for cyclic and unbounded graphs Self-Organizing Maps for cyclic and unbounded graphs M. Hagenbuchner 1, A. Sperduti 2, A.C. Tsoi 3 1- University of Wollongong, Wollongong, Australia. 2- University of Padova, Padova, Italy. 3- Hong Kong

More information

DDoS Detection in SDN Switches using Support Vector Machine Classifier

DDoS Detection in SDN Switches using Support Vector Machine Classifier Joint International Mechanical, Electronic and Information Technology Conference (JIMET 2015) DDoS Detection in SDN Switches using Support Vector Machine Classifier Xue Li1, a *, Dongming Yuan2,b, Hefei

More information

Coupling of surface roughness to the performance of computer-generated holograms

Coupling of surface roughness to the performance of computer-generated holograms Coupling of surface roughness to the performance of computer-generated holograms Ping Zhou* and Jim Burge College of Optical Sciences, University of Arizona, Tucson, Arizona 85721, USA *Corresponding author:

More information

Production of Video Images by Computer Controlled Cameras and Its Application to TV Conference System

Production of Video Images by Computer Controlled Cameras and Its Application to TV Conference System Proc. of IEEE Conference on Computer Vision and Pattern Recognition, vol.2, II-131 II-137, Dec. 2001. Production of Video Images by Computer Controlled Cameras and Its Application to TV Conference System

More information

This paper is part of the following report: UNCLASSIFIED

This paper is part of the following report: UNCLASSIFIED UNCLASSIFIED Defense Technical Information Center Compilation Part Notice ADPO 11846 TITLE: Stream Cipher Based on Pseudo-Random Number Generation Using Optical Affine Transformation DISTRIBUTION: Approved

More information

Relation Organization of SOM Initial Map by Improved Node Exchange

Relation Organization of SOM Initial Map by Improved Node Exchange JOURNAL OF COMPUTERS, VOL. 3, NO. 9, SEPTEMBER 2008 77 Relation Organization of SOM Initial Map by Improved Node Echange MIYOSHI Tsutomu Department of Information and Electronics, Tottori University, Tottori,

More information

Neural Network Based Vision System for Micro Workpieces Manufacturing

Neural Network Based Vision System for Micro Workpieces Manufacturing Neural Network Based Vision System for Micro Workpieces Manufacturing BAIDYK T., KUSSUL E. Center of Applied Science and Technological Development, National Autonomous University of Mexico (UNAM), Cd.

More information

SOMSN: An Effective Self Organizing Map for Clustering of Social Networks

SOMSN: An Effective Self Organizing Map for Clustering of Social Networks SOMSN: An Effective Self Organizing Map for Clustering of Social Networks Fatemeh Ghaemmaghami Research Scholar, CSE and IT Dept. Shiraz University, Shiraz, Iran Reza Manouchehri Sarhadi Research Scholar,

More information

COMBINED METHOD TO VISUALISE AND REDUCE DIMENSIONALITY OF THE FINANCIAL DATA SETS

COMBINED METHOD TO VISUALISE AND REDUCE DIMENSIONALITY OF THE FINANCIAL DATA SETS COMBINED METHOD TO VISUALISE AND REDUCE DIMENSIONALITY OF THE FINANCIAL DATA SETS Toomas Kirt Supervisor: Leo Võhandu Tallinn Technical University Toomas.Kirt@mail.ee Abstract: Key words: For the visualisation

More information

Simulation of WSN in NetSim Clustering using Self-Organizing Map Neural Network

Simulation of WSN in NetSim Clustering using Self-Organizing Map Neural Network Simulation of WSN in NetSim Clustering using Self-Organizing Map Neural Network Software Recommended: NetSim Standard v11.0, Visual Studio 2015/2017, MATLAB 2016a Project Download Link: https://github.com/netsim-tetcos/wsn_som_optimization_v11.0/archive/master.zip

More information

Self-Organized Similarity based Kernel Fuzzy Clustering Model and Its Applications

Self-Organized Similarity based Kernel Fuzzy Clustering Model and Its Applications Fifth International Workshop on Computational Intelligence & Applications IEEE SMC Hiroshima Chapter, Hiroshima University, Japan, November 10, 11 & 12, 2009 Self-Organized Similarity based Kernel Fuzzy

More information

Performance Analysis of various classifiers using Benchmark Datasets in Weka tools

Performance Analysis of various classifiers using Benchmark Datasets in Weka tools Performance Analysis of various classifiers using Benchmark Datasets in Weka tools Abstract Intrusion occurs in the network due to redundant and irrelevant data that cause problem in network traffic classification.

More information

Intrusion Detection by Combining and Clustering Diverse Monitor Data

Intrusion Detection by Combining and Clustering Diverse Monitor Data Intrusion Detection by Combining and Clustering Diverse Monitor Data TSS/ACC Seminar April 5, 26 Atul Bohara and Uttam Thakore PI: Bill Sanders Outline Motivation Overview of the approach Feature extraction

More information

11/14/2010 Intelligent Systems and Soft Computing 1

11/14/2010 Intelligent Systems and Soft Computing 1 Lecture 8 Artificial neural networks: Unsupervised learning Introduction Hebbian learning Generalised Hebbian learning algorithm Competitive learning Self-organising computational map: Kohonen network

More information

Computer Vision 2. SS 18 Dr. Benjamin Guthier Professur für Bildverarbeitung. Computer Vision 2 Dr. Benjamin Guthier

Computer Vision 2. SS 18 Dr. Benjamin Guthier Professur für Bildverarbeitung. Computer Vision 2 Dr. Benjamin Guthier Computer Vision 2 SS 18 Dr. Benjamin Guthier Professur für Bildverarbeitung Computer Vision 2 Dr. Benjamin Guthier 1. IMAGE PROCESSING Computer Vision 2 Dr. Benjamin Guthier Content of this Chapter Non-linear

More information

A Study on Clustering Method by Self-Organizing Map and Information Criteria

A Study on Clustering Method by Self-Organizing Map and Information Criteria A Study on Clustering Method by Self-Organizing Map and Information Criteria Satoru Kato, Tadashi Horiuchi,andYoshioItoh Matsue College of Technology, 4-4 Nishi-ikuma, Matsue, Shimane 90-88, JAPAN, kato@matsue-ct.ac.jp

More information

Parallel string matching for image matching with prime method

Parallel string matching for image matching with prime method International Journal of Engineering Research and Development e-issn: 2278-067X, p-issn: 2278-800X, www.ijerd.com Volume 10, Issue 6 (June 2014), PP.42-46 Chinta Someswara Rao 1, 1 Assistant Professor,

More information

CHAPTER 3 TUMOR DETECTION BASED ON NEURO-FUZZY TECHNIQUE

CHAPTER 3 TUMOR DETECTION BASED ON NEURO-FUZZY TECHNIQUE 32 CHAPTER 3 TUMOR DETECTION BASED ON NEURO-FUZZY TECHNIQUE 3.1 INTRODUCTION In this chapter we present the real time implementation of an artificial neural network based on fuzzy segmentation process

More information

Simple Spatial Domain Filtering

Simple Spatial Domain Filtering Simple Spatial Domain Filtering Binary Filters Non-phase-preserving Fourier transform planes Simple phase-step filters (for phase-contrast imaging) Amplitude inverse filters, related to apodization Contrast

More information

Time Series Prediction as a Problem of Missing Values: Application to ESTSP2007 and NN3 Competition Benchmarks

Time Series Prediction as a Problem of Missing Values: Application to ESTSP2007 and NN3 Competition Benchmarks Series Prediction as a Problem of Missing Values: Application to ESTSP7 and NN3 Competition Benchmarks Antti Sorjamaa and Amaury Lendasse Abstract In this paper, time series prediction is considered as

More information

Fourier analysis of low-resolution satellite images of cloud

Fourier analysis of low-resolution satellite images of cloud New Zealand Journal of Geology and Geophysics, 1991, Vol. 34: 549-553 0028-8306/91/3404-0549 $2.50/0 Crown copyright 1991 549 Note Fourier analysis of low-resolution satellite images of cloud S. G. BRADLEY

More information

Radial Basis Function (RBF) Neural Networks Based on the Triple Modular Redundancy Technology (TMR)

Radial Basis Function (RBF) Neural Networks Based on the Triple Modular Redundancy Technology (TMR) Radial Basis Function (RBF) Neural Networks Based on the Triple Modular Redundancy Technology (TMR) Yaobin Qin qinxx143@umn.edu Supervisor: Pro.lilja Department of Electrical and Computer Engineering Abstract

More information

Unsupervised Learning

Unsupervised Learning Networks for Pattern Recognition, 2014 Networks for Single Linkage K-Means Soft DBSCAN PCA Networks for Kohonen Maps Linear Vector Quantization Networks for Problems/Approaches in Machine Learning Supervised

More information

Smooth shading of specular surfac Title ased high-definition CGH Author(s) MATSUSHIMA, Kyoji 2011 3DTV Conference: The True Vi Citation, Transmission and Display of 3D ) Issue Date 2011 URL http://hdl.handle.net/10112/5575

More information

2. LITERATURE REVIEW

2. LITERATURE REVIEW 2. LITERATURE REVIEW CBIR has come long way before 1990 and very little papers have been published at that time, however the number of papers published since 1997 is increasing. There are many CBIR algorithms

More information

Extract an Essential Skeleton of a Character as a Graph from a Character Image

Extract an Essential Skeleton of a Character as a Graph from a Character Image Extract an Essential Skeleton of a Character as a Graph from a Character Image Kazuhisa Fujita University of Electro-Communications 1-5-1 Chofugaoka, Chofu, Tokyo, 182-8585 Japan k-z@nerve.pc.uec.ac.jp

More information

Supplementary Figure 1: Schematic of the nanorod-scattered wave along the +z. direction.

Supplementary Figure 1: Schematic of the nanorod-scattered wave along the +z. direction. Supplementary Figure 1: Schematic of the nanorod-scattered wave along the +z direction. Supplementary Figure 2: The nanorod functions as a half-wave plate. The fast axis of the waveplate is parallel to

More information

Reverberation design based on acoustic parameters for reflective audio-spot system with parametric and dynamic loudspeaker

Reverberation design based on acoustic parameters for reflective audio-spot system with parametric and dynamic loudspeaker PROCEEDINGS of the 22 nd International Congress on Acoustics Signal Processing Acoustics: Paper ICA 2016-310 Reverberation design based on acoustic parameters for reflective audio-spot system with parametric

More information

Data Mining. Kohonen Networks. Data Mining Course: Sharif University of Technology 1

Data Mining. Kohonen Networks. Data Mining Course: Sharif University of Technology 1 Data Mining Kohonen Networks Data Mining Course: Sharif University of Technology 1 Self-Organizing Maps Kohonen Networks developed in 198 by Tuevo Kohonen Initially applied to image and sound analysis

More information

Surface and thickness profile measurement of a transparent film by three-wavelength vertical scanning interferometry

Surface and thickness profile measurement of a transparent film by three-wavelength vertical scanning interferometry Surface and thickness profile measurement of a transparent film by three-wavelength vertical scanning interferometry Katsuichi Kitagawa Toray Engineering Co. Ltd., 1-1-45 Oe, Otsu 50-141, Japan Corresponding

More information

Novel Method For Low-Rate Ddos Attack Detection

Novel Method For Low-Rate Ddos Attack Detection Journal of Physics: Conference Series PAPER OPEN ACCESS Novel Method For Low-Rate Ddos Attack Detection To cite this article: A A Chistokhodova and I D Sidorov 2018 J. Phys.: Conf. Ser. 1015 032024 View

More information

Segmentation and Object Detection with Gabor Filters and Cumulative Histograms

Segmentation and Object Detection with Gabor Filters and Cumulative Histograms Segmentation and Object Detection with Gabor Filters and Cumulative Histograms Tadayoshi SHIOYAMA, Haiyuan WU and Shigetomo MITANI Department of Mechanical and System Engineering Kyoto Institute of Technology

More information

Robust Steganography Using Texture Synthesis

Robust Steganography Using Texture Synthesis Robust Steganography Using Texture Synthesis Zhenxing Qian 1, Hang Zhou 2, Weiming Zhang 2, Xinpeng Zhang 1 1. School of Communication and Information Engineering, Shanghai University, Shanghai, 200444,

More information

Experimental Competition. Sample Solution

Experimental Competition. Sample Solution The 37th International Physics Olympiad Singapore Experimental Competition Wednesday, 1 July, 006 Sample Solution a. A sketch of the experimental setup Part 1 Receiver Rotating table Goniometer Fixed arm

More information

Images Reconstruction using an iterative SOM based algorithm.

Images Reconstruction using an iterative SOM based algorithm. Images Reconstruction using an iterative SOM based algorithm. M.Jouini 1, S.Thiria 2 and M.Crépon 3 * 1- LOCEAN, MMSA team, CNAM University, Paris, France 2- LOCEAN, MMSA team, UVSQ University Paris, France

More information

A Dendrogram. Bioinformatics (Lec 17)

A Dendrogram. Bioinformatics (Lec 17) A Dendrogram 3/15/05 1 Hierarchical Clustering [Johnson, SC, 1967] Given n points in R d, compute the distance between every pair of points While (not done) Pick closest pair of points s i and s j and

More information

PRELIMINARY RESULTS ON REAL-TIME 3D FEATURE-BASED TRACKER 1. We present some preliminary results on a system for tracking 3D motion using

PRELIMINARY RESULTS ON REAL-TIME 3D FEATURE-BASED TRACKER 1. We present some preliminary results on a system for tracking 3D motion using PRELIMINARY RESULTS ON REAL-TIME 3D FEATURE-BASED TRACKER 1 Tak-keung CHENG derek@cs.mu.oz.au Leslie KITCHEN ljk@cs.mu.oz.au Computer Vision and Pattern Recognition Laboratory, Department of Computer Science,

More information

Local Image Registration: An Adaptive Filtering Framework

Local Image Registration: An Adaptive Filtering Framework Local Image Registration: An Adaptive Filtering Framework Gulcin Caner a,a.murattekalp a,b, Gaurav Sharma a and Wendi Heinzelman a a Electrical and Computer Engineering Dept.,University of Rochester, Rochester,

More information

Video Inter-frame Forgery Identification Based on Optical Flow Consistency

Video Inter-frame Forgery Identification Based on Optical Flow Consistency Sensors & Transducers 24 by IFSA Publishing, S. L. http://www.sensorsportal.com Video Inter-frame Forgery Identification Based on Optical Flow Consistency Qi Wang, Zhaohong Li, Zhenzhen Zhang, Qinglong

More information

A Modular k-nearest Neighbor Classification Method for Massively Parallel Text Categorization

A Modular k-nearest Neighbor Classification Method for Massively Parallel Text Categorization A Modular k-nearest Neighbor Classification Method for Massively Parallel Text Categorization Hai Zhao and Bao-Liang Lu Department of Computer Science and Engineering, Shanghai Jiao Tong University, 1954

More information

Artificial Neural Networks Unsupervised learning: SOM

Artificial Neural Networks Unsupervised learning: SOM Artificial Neural Networks Unsupervised learning: SOM 01001110 01100101 01110101 01110010 01101111 01101110 01101111 01110110 01100001 00100000 01110011 01101011 01110101 01110000 01101001 01101110 01100001

More information

Chapter 7: Competitive learning, clustering, and self-organizing maps

Chapter 7: Competitive learning, clustering, and self-organizing maps Chapter 7: Competitive learning, clustering, and self-organizing maps António R. C. Paiva EEL 6814 Spring 2008 Outline Competitive learning Clustering Self-Organizing Maps What is competition in neural

More information

A study of the Graphical User Interfaces for Biometric Authentication System

A study of the Graphical User Interfaces for Biometric Authentication System A study of the Graphical User Interfaces for Biometric Authentication System Hiroshi Dozono 1, Takayuki Inoue 1, Masanori Nakakun 2 i 1 Faculty of Science and Engineering, Saga University, 1-Honjyo Saga,

More information

Seismic regionalization based on an artificial neural network

Seismic regionalization based on an artificial neural network Seismic regionalization based on an artificial neural network *Jaime García-Pérez 1) and René Riaño 2) 1), 2) Instituto de Ingeniería, UNAM, CU, Coyoacán, México D.F., 014510, Mexico 1) jgap@pumas.ii.unam.mx

More information

Experiment 8 Wave Optics

Experiment 8 Wave Optics Physics 263 Experiment 8 Wave Optics In this laboratory, we will perform two experiments on wave optics. 1 Double Slit Interference In two-slit interference, light falls on an opaque screen with two closely

More information

Fresnel and Fourier digital holography architectures: a comparison.

Fresnel and Fourier digital holography architectures: a comparison. Fresnel and Fourier digital holography architectures: a comparison. Damien P., David S. Monaghan, Nitesh Pandey, Bryan M. Hennelly. Department of Computer Science, National University of Ireland, Maynooth,

More information

Zero Order Correction of Shift-multiplexed Computer Generated Fourier Holograms Recorded in Incoherent Projection Scheme

Zero Order Correction of Shift-multiplexed Computer Generated Fourier Holograms Recorded in Incoherent Projection Scheme VII International Conference on Photonics and Information Optics Volume 2018 Conference Paper Zero Order Correction of Shift-multiplexed Computer Generated Fourier Holograms Recorded in Incoherent Projection

More information

Stability Assessment of Electric Power Systems using Growing Neural Gas and Self-Organizing Maps

Stability Assessment of Electric Power Systems using Growing Neural Gas and Self-Organizing Maps Stability Assessment of Electric Power Systems using Growing Gas and Self-Organizing Maps Christian Rehtanz, Carsten Leder University of Dortmund, 44221 Dortmund, Germany Abstract. Liberalized competitive

More information

Master PE. Bottleneck. Parallelmatching. Network hub. module. Slave PE #2. Slave PE #3. Slave PE #4. Slave PE #1 PE #1 PE #2 PE #3 PE #4.

Master PE. Bottleneck. Parallelmatching. Network hub. module. Slave PE #2. Slave PE #3. Slave PE #4. Slave PE #1 PE #1 PE #2 PE #3 PE #4. Architecture description and prototype demonstration of optoelectronic parallel-matching architecture Keiichiro Kagawa, Kouichi Nitta, Yusuke Ogura, Jun Tanida, and Yoshiki Ichioka?? Department of Material

More information

Fourier, Fresnel and Image CGHs of three-dimensional objects observed from many different projections

Fourier, Fresnel and Image CGHs of three-dimensional objects observed from many different projections Fourier, Fresnel and Image CGHs of three-dimensional objects observed from many different projections David Abookasis and Joseph Rosen Ben-Gurion University of the Negev Department of Electrical and Computer

More information

Machine Learning : Clustering, Self-Organizing Maps

Machine Learning : Clustering, Self-Organizing Maps Machine Learning Clustering, Self-Organizing Maps 12/12/2013 Machine Learning : Clustering, Self-Organizing Maps Clustering The task: partition a set of objects into meaningful subsets (clusters). The

More information

Channel Performance Improvement through FF and RBF Neural Network based Equalization

Channel Performance Improvement through FF and RBF Neural Network based Equalization Channel Performance Improvement through FF and RBF Neural Network based Equalization Manish Mahajan 1, Deepak Pancholi 2, A.C. Tiwari 3 Research Scholar 1, Asst. Professor 2, Professor 3 Lakshmi Narain

More information

Natural Viewing 3D Display

Natural Viewing 3D Display We will introduce a new category of Collaboration Projects, which will highlight DoCoMo s joint research activities with universities and other companies. DoCoMo carries out R&D to build up mobile communication,

More information

Modification of the Growing Neural Gas Algorithm for Cluster Analysis

Modification of the Growing Neural Gas Algorithm for Cluster Analysis Modification of the Growing Neural Gas Algorithm for Cluster Analysis Fernando Canales and Max Chacón Universidad de Santiago de Chile; Depto. de Ingeniería Informática, Avda. Ecuador No 3659 - PoBox 10233;

More information

Overview of nicter - R&D project against Cyber Attacks in Japan -

Overview of nicter - R&D project against Cyber Attacks in Japan - Overview of nicter - R&D project against Cyber Attacks in Japan - Daisuke INOUE Cybersecurity Laboratory Network Security Research Institute (NSRI) National Institute of Information and Communications

More information

A Hierarchial Model for Visual Perception

A Hierarchial Model for Visual Perception A Hierarchial Model for Visual Perception Bolei Zhou 1 and Liqing Zhang 2 1 MOE-Microsoft Laboratory for Intelligent Computing and Intelligent Systems, and Department of Biomedical Engineering, Shanghai

More information

A NEW ALGORITHM FOR OPTIMIZING THE SELF- ORGANIZING MAP

A NEW ALGORITHM FOR OPTIMIZING THE SELF- ORGANIZING MAP A NEW ALGORITHM FOR OPTIMIZING THE SELF- ORGANIZING MAP BEN-HDECH Adil, GHANOU Youssef, EL QADI Abderrahim Team TIM, High School of Technology, Moulay Ismail University, Meknes, Morocco E-mail: adilbenhdech@gmail.com,

More information

Multiresolution Texture Analysis of Surface Reflection Images

Multiresolution Texture Analysis of Surface Reflection Images Multiresolution Texture Analysis of Surface Reflection Images Leena Lepistö, Iivari Kunttu, Jorma Autio, and Ari Visa Tampere University of Technology, Institute of Signal Processing P.O. Box 553, FIN-330

More information

Local optimization strategies to escape from poor local minima

Local optimization strategies to escape from poor local minima Header for SPIE use Local optimization strategies to escape from poor local minima Florian Bociort, lexander Serebriakov and Joseph Braat Optics Research Group, Delft University of Technology Lorentzweg

More information

Fuzzy-Kernel Learning Vector Quantization

Fuzzy-Kernel Learning Vector Quantization Fuzzy-Kernel Learning Vector Quantization Daoqiang Zhang 1, Songcan Chen 1 and Zhi-Hua Zhou 2 1 Department of Computer Science and Engineering Nanjing University of Aeronautics and Astronautics Nanjing

More information

13. Learning Ballistic Movementsof a Robot Arm 212

13. Learning Ballistic Movementsof a Robot Arm 212 13. Learning Ballistic Movementsof a Robot Arm 212 13. LEARNING BALLISTIC MOVEMENTS OF A ROBOT ARM 13.1 Problem and Model Approach After a sufficiently long training phase, the network described in the

More information

DDoS Attack Detection Using Moment in Statistics with Discriminant Analysis

DDoS Attack Detection Using Moment in Statistics with Discriminant Analysis DDoS Attack Detection Using Moment in Statistics with Discriminant Analysis Pradit Pitaksathienkul 1 and Pongpisit Wuttidittachotti 2 King Mongkut s University of Technology North Bangkok, Thailand 1 praditp9@gmail.com

More information

Robust and Accurate Detection of Object Orientation and ID without Color Segmentation

Robust and Accurate Detection of Object Orientation and ID without Color Segmentation 0 Robust and Accurate Detection of Object Orientation and ID without Color Segmentation Hironobu Fujiyoshi, Tomoyuki Nagahashi and Shoichi Shimizu Chubu University Japan Open Access Database www.i-techonline.com

More information

Neuro-Fuzzy Comp. Ch. 8 May 12, 2005

Neuro-Fuzzy Comp. Ch. 8 May 12, 2005 Neuro-Fuzzy Comp. Ch. 8 May, 8 Self-Organizing Feature Maps Self-Organizing Feature Maps (SOFM or SOM) also known as Kohonen maps or topographic maps were first introduced by von der Malsburg (97) and

More information

CHAPTER 6 MODIFIED FUZZY TECHNIQUES BASED IMAGE SEGMENTATION

CHAPTER 6 MODIFIED FUZZY TECHNIQUES BASED IMAGE SEGMENTATION CHAPTER 6 MODIFIED FUZZY TECHNIQUES BASED IMAGE SEGMENTATION 6.1 INTRODUCTION Fuzzy logic based computational techniques are becoming increasingly important in the medical image analysis arena. The significant

More information

Fingerprint Recognition using Texture Features

Fingerprint Recognition using Texture Features Fingerprint Recognition using Texture Features Manidipa Saha, Jyotismita Chaki, Ranjan Parekh,, School of Education Technology, Jadavpur University, Kolkata, India Abstract: This paper proposes an efficient

More information

5.6 Self-organizing maps (SOM) [Book, Sect. 10.3]

5.6 Self-organizing maps (SOM) [Book, Sect. 10.3] Ch.5 Classification and Clustering 5.6 Self-organizing maps (SOM) [Book, Sect. 10.3] The self-organizing map (SOM) method, introduced by Kohonen (1982, 2001), approximates a dataset in multidimensional

More information

Anomaly Detection System for Video Data Using Machine Learning

Anomaly Detection System for Video Data Using Machine Learning Anomaly Detection System for Video Data Using Machine Learning Tadashi Ogino Abstract We are developing an anomaly detection system for video data that uses machine learning. The proposed system has two

More information

Shortening Time Required for Adaptive Structural Learning Method of Deep Belief Network with Multi-Modal Data Arrangement

Shortening Time Required for Adaptive Structural Learning Method of Deep Belief Network with Multi-Modal Data Arrangement Shortening Time Required for Adaptive Structural Learning Method of Deep Belief Network with Multi-Modal Data Arrangement arxiv:180703952v1 [csne] 11 Jul 2018 Shin Kamada Graduate School of Information

More information

Fountain Codes Based on Zigzag Decodable Coding

Fountain Codes Based on Zigzag Decodable Coding Fountain Codes Based on Zigzag Decodable Coding Takayuki Nozaki Kanagawa University, JAPAN Email: nozaki@kanagawa-u.ac.jp Abstract Fountain codes based on non-binary low-density parity-check (LDPC) codes

More information

Image-Space-Parallel Direct Volume Rendering on a Cluster of PCs

Image-Space-Parallel Direct Volume Rendering on a Cluster of PCs Image-Space-Parallel Direct Volume Rendering on a Cluster of PCs B. Barla Cambazoglu and Cevdet Aykanat Bilkent University, Department of Computer Engineering, 06800, Ankara, Turkey {berkant,aykanat}@cs.bilkent.edu.tr

More information

/00/$10.00 (C) 2000 IEEE

/00/$10.00 (C) 2000 IEEE A SOM based cluster visualization and its application for false coloring Johan Himberg Helsinki University of Technology Laboratory of Computer and Information Science P.O. Box 54, FIN-215 HUT, Finland

More information

Image Restoration by Revised Bayesian-Based Iterative Method

Image Restoration by Revised Bayesian-Based Iterative Method ADVCOMP 2011 : The Fifth International Conference on Advanced Engineering Computing and Applications in Sciences Image Restoration by Revised Bayesian-Based Iterative Method Sigeru Omatu, Hideo Araki Osaka

More information

How to Predict Viruses Under Uncertainty

How to Predict  Viruses Under Uncertainty How to Predict Email Viruses Under Uncertainty InSeon Yoo and Ulrich Ultes-Nitsche Department of Informatics, University of Fribourg, Chemin du Musee 3, Fribourg, CH-1700, Switzerland. phone: +41 (0)26

More information

Shading of a computer-generated hologram by zone plate modulation

Shading of a computer-generated hologram by zone plate modulation Shading of a computer-generated hologram by zone plate modulation Takayuki Kurihara * and Yasuhiro Takaki Institute of Engineering, Tokyo University of Agriculture and Technology, 2-24-16 Naka-cho, Koganei,Tokyo

More information

Reduction of reconstructed particle elongation using iterative min-max filtering in holographic particle image velocimetry

Reduction of reconstructed particle elongation using iterative min-max filtering in holographic particle image velocimetry Reduction of reconstructed particle elongation using iterative min-max filtering in holographic particle image velocimetry Yohsuke Tanaka 1, *, Shigeru Murata 1 1: Department of Mechanical System Engineering,

More information

3D Visualization of Sound Fields Perceived by an Acoustic Camera

3D Visualization of Sound Fields Perceived by an Acoustic Camera BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 15, No 7 Special Issue on Information Fusion Sofia 215 Print ISSN: 1311-972; Online ISSN: 1314-481 DOI: 1515/cait-215-88 3D

More information

Machine Learning Based Autonomous Network Flow Identifying Method

Machine Learning Based Autonomous Network Flow Identifying Method Machine Learning Based Autonomous Network Flow Identifying Method Hongbo Shi 1,3, Tomoki Hamagami 1,3, and Haoyuan Xu 2,3 1 Division of Physics, Electrical and Computer Engineering, Graduate School of

More information

Lens Design I. Lecture 11: Imaging Herbert Gross. Summer term

Lens Design I. Lecture 11: Imaging Herbert Gross. Summer term Lens Design I Lecture 11: Imaging 2015-06-29 Herbert Gross Summer term 2015 www.iap.uni-jena.de 2 Preliminary Schedule 1 13.04. Basics 2 20.04. Properties of optical systrems I 3 27.05. 4 04.05. Properties

More information

Feature Based Watermarking Algorithm by Adopting Arnold Transform

Feature Based Watermarking Algorithm by Adopting Arnold Transform Feature Based Watermarking Algorithm by Adopting Arnold Transform S.S. Sujatha 1 and M. Mohamed Sathik 2 1 Assistant Professor in Computer Science, S.T. Hindu College, Nagercoil, Tamilnadu, India 2 Associate

More information

Mineral Exploation Using Neural Netowrks

Mineral Exploation Using Neural Netowrks ABSTRACT I S S N 2277-3061 Mineral Exploation Using Neural Netowrks Aysar A. Abdulrahman University of Sulaimani, Computer Science, Kurdistan Region of Iraq aysser.abdulrahman@univsul.edu.iq Establishing

More information

Synthesis of Facial Images with Foundation Make-Up

Synthesis of Facial Images with Foundation Make-Up Synthesis of Facial Images with Foundation Make-Up Motonori Doi 1,RieOhtsuki 2,RieHikima 2, Osamu Tanno 2, and Shoji Tominaga 3 1 Osaka Electro-Communication University, Osaka, Japan 2 Kanebo COSMETICS

More information

Facial expression recognition using shape and texture information

Facial expression recognition using shape and texture information 1 Facial expression recognition using shape and texture information I. Kotsia 1 and I. Pitas 1 Aristotle University of Thessaloniki pitas@aiia.csd.auth.gr Department of Informatics Box 451 54124 Thessaloniki,

More information

Lecture Topic Projects

Lecture Topic Projects Lecture Topic Projects 1 Intro, schedule, and logistics 2 Applications of visual analytics, basic tasks, data types 3 Introduction to D3, basic vis techniques for non-spatial data Project #1 out 4 Data

More information

DESIGN OF KOHONEN SELF-ORGANIZING MAP WITH REDUCED STRUCTURE

DESIGN OF KOHONEN SELF-ORGANIZING MAP WITH REDUCED STRUCTURE DESIGN OF KOHONEN SELF-ORGANIZING MAP WITH REDUCED STRUCTURE S. Kajan, M. Lajtman Institute of Control and Industrial Informatics, Faculty of Electrical Engineering and Information Technology, Slovak University

More information

What is a receptive field? Why a sensory neuron has such particular RF How a RF was developed?

What is a receptive field? Why a sensory neuron has such particular RF How a RF was developed? What is a receptive field? Why a sensory neuron has such particular RF How a RF was developed? x 1 x 2 x 3 y f w 1 w 2 w 3 T x y = f (wx i i T ) i y x 1 x 2 x 3 = = E (y y) (y f( wx T)) 2 2 o o i i i

More information