2014

Browse

Filters

201424

Settings

Search Results

Now showing 1 - 10 of 24
  • Thumbnail Image
    Item
    Candidate Gene Prioritization Using Unique Pattern Indexing and Mapping Techniques
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Mutakabbir, Kazi Mahbub; Mahin, Shah S
    “Prioritizing the candidate gene is amongst the notable work in bioinformatics. Techniques have been applied to reduce the number of promising genes for a certain disease. Previous works were done by using PageRank and HITS algorithm on graph based network. However using frequent pattern mining this prioritizing can be made more efficient. In this paper, we propose four algorithms. The first one indexes the unique sequences of length four using an integer value. The second algorithm finds the frequency of the frequent patterns of various lengths by searching through the integer values instead of the patterns themselves. Third one weights the candidate gene in compare with the genes of database. Fourth algorithm creates the graph network and ranks the candidate gene. All this is done highly efficiently by the use of mapping techniques e.g. HashMap. Due to its highly frugal nature, the proposed algorithm can reduce typical memory usage by 37.5% at the very minimum.”
  • Thumbnail Image
    Item
    A subset selection method using Filter and wrapper algorithms based on the nature of the expression values in microarray data sets for gene feature selection
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Azad, Tamzid; Islam, Md. Mazharul
    In the field of micro-array data analysis the crucial first step is gene selection. The process refers to selecting a subset consisting of a few genes which are of genetic significance out of thousands of genes to make the job of the classifier algorithm computationally easy and efficient at the same time. Feature selection plays an important role in classification. The first set of data are gene expression profiles from Acute Lymphoblastic Leukemia (ALL) patients. In this paper an algorithm is proposed for feature subset selection (FFS) which is based on the nature of intensity values in microarray datasets. The proposed method is a combination of a filter and a wrapper algorithm which selects subsets. It is based on two assumptions. Our results demonstrate the importance of feature selection in accurately classifying new samples
  • Thumbnail Image
    Item
    Hybrid Gene Selection Framework using Adaptive Wrapper and Filtering Techniques
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-10-15) Islam, Md Anisul; Mottalib, Md Mozaharul
    Analysing the thousands of gene expression values is a difficult task due to the curse of dimensionality of data produced by Microarray chips. Primary role of an effective feature selection model is to simplify this task. To simplify the task of disease classification and predicting cancer, feature selection plays a vital role through removing less informative genes. In this study, we propose a hybrid approach to gene selection using adaptive filter and adaptive wrapper approach. As filter method exhibits some limitations, an adaptive form of filtering has been employed that iteratively selects genes in each iteration and emphasizes on the misclassified samples and in subsequent iteration tries to find out the effective genes for misclassified samples. This approach performs better than traditional filter methods as it focuses on its weaknesses. In gene selection, Artificial Neural Networks (ANN) are mostly used as a classifier. In this study, adaptive ANN has been used as an internal wrapper. This helps to generate a better subset of genes. The proposed hybrid approach is applied on leukaemia, colon and lung cancer benchmarked datasets. Better result has been found compared to other well-known approaches.
  • Thumbnail Image
    Item
    An Optimized Algorithm to Find Maximum Parsimonious Tree Using PrimeNucleotide Based Approach
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Ajwad, Rasif; Hossain, Syed Nayem
    Molecular phylogeny based on the nucleotide or amino acid sequence comparison has become a widespread tool for general taxonomy and evolutionary analysis. Molecular phylogeny methods are often free of problems which arise while applying phenotypic phylogeny. So, we prefer phylogenetic methods for classifying the organisms in any evolutionary situation. Phylogenetic inference methods like Maximum parsimony perform exhaustive search strategy to extract evolutionary information from genomic sequences. It is a simple but popular technique used in cladistics to infer a phylogenetic tree for a set of taxa (commonly of species or reproductively isolated populations of a single species) on the basis of some observed data on the similarities and differences among taxa. The relationships among organisms or genes are studied by comparing the homologues of DNA and protein sequences. However, complexity arises when we increase the number of sequences involved, as the number of possible solutions increase exponentially alongside. In our paper, we have proposed an algorithm which identifies the highest repeating nucleotide (PrimeNucleotide) from the informative site efficiently to fix one ParentNode with the best fitted nucleotide using a predefined WeightMatrix to find the most parsimonious phylogenetic tree in linear time. The algorithm has been applied on the genome sequences of different bacteria and viruses to ensure its efficiency and universality. The results obtained were similar to the traditional Transverse parsimony method and a significant improvement in both time consumption and memory usage rate were achieved.
  • Thumbnail Image
    Item
    Human Activity Recognition using Dynamic Time Warping
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Hasan, G. M. Mukit; Iftekhar, Kazi MD
    Human activity recognition is now a well-known field of Human Computer Interaction (HCI) because its capability of providing personalized support using different applications. The recognition of human activities has become a task of high interest within the field, especially for medical, military, and security applications. Its applications include surveillance systems, patient monitoring systems, and a variety of systems that involve interactions between persons and electronic devices such as human-computer interfaces. In our work we will try to recognize human activity using wearable sensors. The works done before had the problem of using multimodal system with satisfactory results. Computation of the inputs to recognize activities are not simple. So, we designed a multi-modal system that will take accelerometer, gyroscope and ultrasonic sensor’s data as input and use optimized Dynamic Time Warping algorithm to classify data to recognize activity with satisfactory success rate.
  • Thumbnail Image
    Item
    Offline Bangla handwritten character recognition
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Islam, Yamin Bin; Rahman, Md. Taiyebur
    Optical character recognition is an attractive subject for research work now days. In our study of this topic for Bangla Character domain, we have come around some methods of feature extraction and classification along with various preprocessing steps. We have exploited these techniques for their advantages and disadvantages. Among these, the methods that came to our attention because of their high accuracy are Stroke Extraction and encoding which is mainly mathematical, Chaincode extraction based on the direction of the points and Principal Component Analysis(PCA) of the character image. In our research, we have combined these methods and implemented the system in Matlab that gave us better accuracy than their individual implementation.
  • Thumbnail Image
    Item
    Fingertip Detection and Finger Identification for Real-time Hand Gesture Recognition using Kinect
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Marouf, Ahmed Al; Shondipon, Shaumic
    Fingertip Detection and finger identification is one of the main challenges of gesture spotting and recognition. The topic is mostly related to the Human computer Interaction(HCI) area. We have proposed a system for detection of fingertip and identification of finger by several steps like pre-processing, processing, hand segmentation, palm point identification, fingertip detection, finger identification. Pre-processing includes the thresholding ,RGB to Grayscale conversion and color & depth calibration. Determining the minimum depth value from the Kinect camera, determining the segmentation threshold, cropping the region of interest and edge detection are the steps of processing. Depth and color segmentation and calibration is done for hand segmentation. For fingertip detection we have merged two existing idea from state of the art in a manner that it minimizes the limitations of both approach. Ellipse fitting technique is applied for palm point determination and point-to-point scanning is done for fingertip locating. For finger identification, we have proposed a new model, 4Y - model based on iterative Hill climbing Algorithm for finger identification. The main contribution in this area would be the new approaches of fingertip detection and finger identification.
  • Thumbnail Image
    Item
    Co-ordination of Streams of Data
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Mamun, Abdullah Al; Hasan, Tahmeed
    Here in this thesis work we have proposed a model for some issues from a new data processing model. In this type of model data we are dealing with does not possess any relation that is persistent, rather this data have some different properties. They arrive in real time, continuous, time varying, rapid and infinite streams of data. This paper introduces a model for the delayed and unordered arrival of streams. In addition, this thesis work also introduces the concept of using buffers and algorithmic issues for stream management purpose reviewing the past related works and also the ongoing research in this area. This thesis proposes a model and an algorithm that will deal with the coordination of the tuples that do not arrive in due time and in some cases they just lost forever. The model introduces some modules that will depict the total work flow and the algorithm will propose some data set that will enable one to choose the tradeoff depending on the system requirements implementing the ordering mechanism.
  • Thumbnail Image
    Item
    Teacher Evaluation System of Islamic University of Technology
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Saleh, Abdullah Al; Bhuiyan, Ishmam Haque
    Automation system is now being used in every aspects of life. It has made our life easier and the possibility of covering greater area of work has increased due to the automation process. Completing track of work and keeping records of work has become much easier because it reduced the huge amount of paper work which was needed before. The systematic automation of virtually every facet of life activity has today become more of a necessity than just mere luxury and the putting of such system online is even becoming increasingly imperative. With this respect the idea to develop an Online Teacher Evaluation System is being conceived. With a total population of over 1200 and each student each student evaluate the teaching of an average of six subjects and corresponding labs, the volume of paper work that has to be done, sorted and listed is enormous. Every semester around 12000 evaluation sheets are collected by the respective department secretaries. Getting to summarize those evaluation sheets is a time consuming and difficult work. So to make this evaluation system fully functional and the whole process easier, the idea of teacher evaluation system has been introduced. Each student is assigned a username and password with which he can access the system and evaluate only the teachers who were in that particular semester. Teachers can also access the system and see their evaluation. Head of the departments and vice chancellor can also access the system with various privileges. Only the VC has unrestricted access across the system. There will also be an admin in the system. He is the one who is responsible for creating new users and their passwords, granting different level of the system according to the user. In a nutshell, if the system is deployed, it will make the evaluation process suitable and easier and reduce the huge amount of paperwork. The system offers very user friendly interface, graphical representation of evaluation and gives an overview of whole evaluation system.
  • Thumbnail Image
    Item
    Implementation of sip over VoIP
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2014-11-15) Rahi, Mohammad Muntashir Are; Fuad, Nafis
    Back in the days of wired telephony, when all phone calls went over the PSTN, businesses would purchase “trunks” – a dedicated line or a bundle of circuits – from their service provider. Today, we have adapted the concept of “trunking” to the IP-enabled landscape resulting in lower telephony costs and rapid return on investment (ROI) plus the opportunity for enhanced communications both within the enterprise and with vendors, customers and partners. A SIP trunk is a service offered by an ITSP to use SIP to set up communications between an enterprise PBX and the ITSP. A trunk includes multiple voice sessions – as many as the enterprise needs. While some see SIP as just voice, SIP trunking can also serve as the starting point for the entire breadth of real time communications possible with the protocol, including Instant Messaging, presence applications, white boarding and application sharing. The potential for a rapid return on investment is a key driver of SIP trunk deployments. However, maximum return on investment can be achieved when you extend VoIP outside of the corporate LAN. In terms of infrastructure purchases, SIP trunks provide an immediate cost-savings. They eliminate the need to purchase costly BRIs, PRIs or PSTN gateways. The productivity benefits with SIP and SIP trunking are also significant. By extending the SIP capabilities of the corporate network outside the LAN, satellite offices, remote workers and even customers can use VoIP and other forms of real-time communications applications to break down barriers of geography to share ideas and increase productivity. There are three components necessary to successfully deploy SIP trunks: a PBX with a SIP-enabled trunk side, an enterprise edge device understanding SIP and an Internet telephony or SIP trunking service provider. Equipment based on the SIP protocol – SIP phones, IP-PBXs etc. – have been around for some time. Now that SIP trunks have gained momentum, it has become 3 | P a g e important to ensure that equipment works together. It is for this reason that standards such as SIP connect™ have become so critical. SIP connect was developed by the SIP Forum as a set of best practices for interfacing an enterprise PBX implementation with an ITSP that attempts to eliminate some of the unknowns and incompatibilities of mixing equipment from different vendors in a single environment. Like any application that opens the network to the Internet, SIP trunking deployments have security considerations, but there are ways to maximize enterprise security. One of the most effective techniques is to address SIP security the same way data security is addressed - at the enterprise edge. SIP server and SIP proxy technologies offer maximum control over the flow of SIP traffic, enabling the administrator to ensure correct routing, apply verification and authentication policies and mitigate Denial-of-Service attacks. Voice quality is not an issue with SIP trunking if proper Quality of Service (QoS) measures are applied, such as over provisioning of links, and prioritization of voice traffic. Reliability is also a moot point. In fact, SIP trunks can be more reliable than the traditional PSTN as a number of failover solutions can be implemented.