Research Interests. Education

Size: px

Start display at page:

Download "Research Interests. Education"

Angelica Bridges
5 years ago
Views:

Personal Information Ali Mollahosseini Date of Birth: 21 September 1984 Mailing address: No. 2, Ehsani Alley, Between Khosh & Ghasredasht, Emam Khomeini Str. Tehran, Iran, 1346849197 Phone No.

1 Personal Information Ali Mollahosseini Date of Birth: 21 September 1984 Mailing address: No. 2, Ehsani Alley, Between Khosh & Ghasredasht, Emam Khomeini Str. Tehran, Iran, Phone No.: Personal webpage: Research Interests Bioinformatics & Health informatics o DNA microarray and SNP dataset classification o DNA and protein sequences and alignment o protein/rna structure and function prediction o bimolecular interaction networks and protein-protein interactions Machine learning and Data Mining o Kernel methods o Solving so called 'small N Large P' problems o simultaneous feature extraction and classification techniques o Fuzzy Database Computational Intelligence o Genetic & Evolutionary Algorithms o Particle Swarm Optimization o Artificial Immune Systems o Fuzzy Systems and Neural Networks Steganalysis techniques using machine learning techniques o Classification of steganographic and cover images Education M.Sc. in Computer Engineering (Artificial Intelligence discipline), Dept. of Computer Engineering, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran. (GPA: 17.01/20 3.5/4.0) B.Sc. in Computer Engineering (Software Engineering discipline), Dept. of Computer Engineering, Iran University of Science and Technology, Tehran, Iran. (GPA: 15.92/ /4.0) (The last two years GPA: 16.50/ /4.0) Mathematics & Physics Diploma, Shahram Emami High school, Tehran, Iran. (GPA: 19.04/ /4.0) 1

2 Research Experience Amir Kabir University (Tehran Polytechnic), Dept. of Computer Engineering, Image Processing and Pattern Recognition Laboratory, 2007-Present: o DNA-microarray gene selection & clustering algorithms. o Population based algorithms (especially in PSO & AIS). o Classification of Steganographic Jpeg Images using Artificial Immune System. Iran University of Science and Technology, Dept. of Computer Engineering, Fuzzy and Neural Network Laboratory, : o Fuzzy Database. o Query Processing Publications 1. Ali Mollahosseini, Sara Sadegh Tabrizi; "Active Learning Method and its application in modeling and fuzzy controllers: A survey"; 7 th Computer Engineering Student Conference (CESC); 2004 Tehran (Poster) 2. Ali Mollahosseini, Omid Ranjbar Pouya; "A deep review of query processing in fuzzy DataBases and their problems"; 4 th international Conference of Computer Engineering at University of Science & Culture; 2006; Tehran. 3. Naser Mozayani, Ali Mollahosseini, Omid Ranjbar Pouya; "An optimal algorithm for calculating nested Query in fuzzy Databases"; 12 th International Conference of CSI; 2007; Tehran. 4. Ali Mollahosseini, Hoesin Amirkhani, Mohammad Rahmati; "A Novel classifier based on Ada-boost for Steganalysis", 15th CSI Computer Conference: Symposium on Intelligent Systems and Soft Computing, 2010; Tehran. Ongoing Research Classification Microarray Data using GO information Classification of Steganographic Jpeg Images Using an Artificial Immune System Teaching Experience Teacher Assistant of Statistical Pattern Recognition 2 semester Teacher Assistant of Theory of Automata and Languages 3 semester Teacher Assistant of Programming Languages 1 semester 2

3 Work Experience Fan Avaran Asre Novin Co., : o Working part time in software developing team as software analyst, designer and developer. MAHFA Co., : o Working full time as software project manager. Fan Avaran Asre Novin Co., 2008-Present: o Working at R&D department of FAN Co Honors & Awards Fist Candidate of PhD erasmus mundus full scholarship for 34 month for University of Algarve, Portugal Ranked 40 th among over 10,000 applicants in Iran s Nationwide Master of Science Entrance Exam for Computer Engineering, Feb Praised by The Minister of Education as the 3 th student of 10 th zone of Tehran. Skills Programming languages: C#.net, ASP.net, Java, C, C++, CUDA. Database design & administration: MySQL, SQL Server, Oracle. Operating Systems: Windows, Linux (Fedora, Obunto) CAD Tools: MATLAB, Rational Rose, Office. Seminars 7 th Computer Engineering Student Conference (CESC); 2004 Tehran 4 th international Conference of Computer Engineering at University of Science & Culture; 2006; Tehran 12 th International Conference of CSI; 2007; Tehran 13 th International Conference of CSI; 2008; Kish Island 14 th International Conference of CSI; 2009; Tehran Extracurricular Involvements Member of Student Council: Member of IUST Graduated Students of Engineering : 2006-present 3

4 Languages Language Reading Writing Speaking Certifications Skill Skill Skill Persian/ Farsi Native Native Native English Good Good Good TOEFL: 95 (R24,L27,S22,W22) GRE: (Q800, V270, AW2.5) German/ Deutsch Intermediate Intermediate Tolerable ZD(ZertifKat Deutsch) score 2, 800 hour learning Germany at Gohte Institute, Tehran B.Sc. Project A review of query processing in fuzzy Databases under supervision of Prof. Naser Mozayani. This project mainly focused on how fuzzy databases process queries (especially nested queries which take much process time rather than relational and other normal databases with crisp data). At this Project we manly work on enhance them by introducing new algorithm that theoretically enhance their time order (Although it use much memory), considering nowadays server having huge memories in order to paging data. Despite it was only a B.Sc. Project, we tried to study this field and its related source without passing any fuzzy course at B.Sc and our conscientious attitude led to publishing two papers in this field (item 2 & 3 conference Papers). M.Sc. Thesis Blind Image Steganalysis Using an Artificial Immune System under supervision of Prof. Mohammad Rahmati. Steganography is the art and science of hidden communication. It attempts to conceal data in a way that no one else would be aware of the existence of the hidden data except the intended receiver. In contrast to steganography, steganalysis focus on detecting, tracking, extracting, and modifying secret messages transmitted through a covert channel. As steganographic techniques become more and more sophisticated, more advanced steganalysis schemes are needed in practice. Nowadays, digital image is very popular and can be obtained almost everywhere. It can tolerate large amount of perceptually redundant information without any visible clue. Therefore, digital image is an ideal cover media for steganography yet a big challenge for steganalysis. In addition, Artificial Immune System approach, which makes its way into the mainstream of detection abnormality soon, solve abnormal detection problems much 4

5 better than using pervious approaches. These systems are inspired by various aspects of the immune systems of mammals. Some of these aspects, such as the distinction between self and non-self and the concept of negative selection, have a natural and intuitive fit for applications involving computer security, network intrusion or abnormality detection. This thesis mainly focuses on proposing new classification algorithm for Image Steganalysis that would use Artificial Immune System Algorithms, which is proposed to blindly (i.e., without knowledge of the steganographic schemes) determine the existence of hidden messages in Jpeg image. From this kind of view, we design a new classification based on combination of AIS and Pattern Recognition techniques especially boosting classification, which distinguish nonself images (steganographic images) form self images (non- steganographic images or cover Images). We also use enhanced AIS using danger theory by studying Immune System of human body. We propose a combination of an enhanced negative selection algorithm and clonal selection, with a boosting classification approach for training weak learners. Several key characteristics make this method a state-of-the-art advance in the previous AIS classification methods. First, individual-specific size of weak learner (same as detectors in NSA, CSA), is utilized to maximize the anomaly coverage. Second, using statistical estimation of each weak learner's coverage area, which make train each new weak learner based on not only True Negative (or False Positive in other approaches), but also coverage of other weak learners in order to minimizing number of detectors. Moreover, because the JPEG format is the most dominant image format for image storage and exchange, several steganographic techniques working on JPEG images have been proposed in recent years. In this thesis, we propose to use third order statistics of both the intrablock and interblock correlations among JPEG coefficients to enhance steganalysis capability. However, because of this approach, our proposed feature set increased exponentially, compared with second order statistics. We also obtain an AIS based method for reducing feature dimensions. This method is similar to using GA as wrapper method for so called 'Large P Small N' problem. However, we utilize aspect of human's Immune Systems. In this approach, we select a random subset of feature set and improve it by Clonal Selection (i.e. hyper mutation and clonal expansion). Finally, test our proposed subset using a validation set, and in case of violation of predefined criteria (Algorithm's parameter), a danger signal (inspired from human's Immune Systems) would generate. This wrapper like approach simultaneously works through training new weak learners. 5

Personal Information. Research Interest. Education. Honors &

Personal Information. Research Interest. Education. Honors & CV B. Hoda Helmi Computer Engineering Dept., Iran University of Science and Technology April, 12 th, 2013, Phone: (+98)2177451896 Cell: (+98)915 509 6783 E mail: hodahelmi@gmail.com Personal Information