SymSCMA – 2016
Permanent URI for this collectionhttp://repository.kln.ac.lk/handle/123456789/15544
Browse
14 results
Search Results
Item New Feature Selection Method for High Dimensional Gene Data(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Fajila, M.N.F.; Nawarathna, R.D.Dimensionality reduction (i.e., feature selection) is an essential technique in data science when handling high dimensional data such as cancer microarray samples. Cancer microarray experiments normally provide a large number of data which is assumed to contain many features, called, genes. However, genes can be either redundant or irrelevant, and thus be removed without incurring much loss of information. A small number of samples with a large number of genes is the major problem in microarray data analysis. In this study, a new machine learning method, namely, hybrid wrapper – filter feature selection is proposed for gene selection. This approach combines the genes selected by both filter and wrapper feature selection methods. Further, it uses a least priority feature elimination procedure where the genes with the lowest priority are eliminated. The propsoed technique is validated and evaluated on two microarray data sets namely, leukemia and colon cancer data sets. With gene selection performed by the proposed method, it helps to classify the leukemia microarray samples with perfect classification (100%) and to classify the colon cancer data set only with two misclassifications giving an accuracy of 90.5%. Results show that the proposed technique is extremely efficient in terms of the computational time too.Item Android Application using Review Mining: Ceylon Tour Guide(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Nishanthan, M.; Kavitha, S.; Mosharani, S.; Amuthapriya, S.; Vithana, V.N.In the 21st century, Sri Lanka is one of the most visited tourist countries in Asia. Though, the rate of tourists visiting Sri Lanka is rapidly increasing, most of these tourists do not have a clear idea of where to visit. Several web sites such as Agoda, Trip Advisor let tourists see comments and rate places. There are very few applications which assess these public reviews and provide a comprehensive rating specially tailored to Sri Lanka. Hence tourists end up spending a lot of time and effort in looking for locations of their interest. To overcome this problem, our system, Ceylon Tour Guide presents a Tour guide application that provides a comprehensive rating of tourist locations in Sri Lanka using text mining techniques. Public reviews from well-established tourist web sites such as Agoda are collected and stored in a database. The stored details are mined with the text mining algorithms. With the aid of algorithms, mined data will be categorized into positive, neutral and negative reviews. Using the Android application user can search and see the ratings of the important locations. After ending 30 years of civil war, the tourism sector has a vast development. This application will help popularize the important locations in Sri Lanka including north and thus, help increase the economic growth of the country.Item An Attempted Crowd and Floor Management System(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Bandara, R.M.C.; de Silva, L.N.B.; Bandara, D.D.M.M.; Silva, R.L.C.; Guruge, I.This research article represents a system which enhances crowd handling and floor management techniques created with the assistance of Closed-Circuit Television (CCTV) and proximity beacons (BLE). With an aid of a desktop application crowd at a particular event/party will be observed with the use of CCTV cameras, which contains a video surveillance system. Video processing will also be used to keep track of a specific number of people roaming around an area and to make sure that no one enters any restricted areas. The visitors who arrive into a particular event will have a mobile application, which would guide them to navigate throughout an entire area and make sure that they do not visit any restricted areas. Bluetooth low energy (BLE) Proximity beacons will be used send signals throughout the generated map containing stalls Via Bluetooth; it also indicates the restricted areas on the map with the aid of the mobile application. This system will be beneficial for both organizers and visitors, organizers will finally find it easy to keep track of people and the visitors will be able to visit the areas they wish to without any difficulty.Item Automated Help Aid for Visually Impaired People using Obstacle Detection and GPS Technology: AKSHI(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Kaushalya, V.S.S.; Premarathne, K.D.D.P.; Shadir, H.M.; Krithika, P.; Fernando, S.G.S.There is an immense population of the visually impaired people around the globe. The motivation of the system is to help blind people where they can move along without a dependent. Since there were existing on blind stick navigation, Akshi composes some unique features. Such as obstacle detection, Pedestrian crossing identification, voice commands detection, outdoor navigation system and a tracking system with the use of Global Positioning system (GPS), Radio Frequency Identification (RFID), Android, and Global system for mobile communication (GSM) Technologies. This research has been developed by user friendly and cost- effective manner with the use of open source software and Hardware devices like Raspberry pi circuit, RFID Reader etc. Specially the research is focused on the safety of the visually impaired people and provide interoperable solutions for the problems that they face in their day-to-day life without any hesitation.Item 2-Tuple Fuzzy Linguistic Model to Evaluate the Risk of Invasive Plant Species(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Peiris, H.O.W.; Perera, S.S.N.; Chakraverty, S.; Ranwala, S.M.W.Management of invasive species can appear to be a complicated and unending task. In order to manage the spread, these species need to be undergone any risk assessment during their introduction. The aim of this study is to evaluate the aggregate risk of Invasive Alien Species (IAS) using invasive attributes. We use the 2-tuple fuzzy linguistic representation to develop the model without loss of information in which occur in ordinary linguistic operators. These risk values are compared with the National Risk assessment scores which are in the form of Linguistic labels. The proposed model is validated using few known noninvasive species in Sri Lanka. The model gives significant predictions and it is found to be a better tracking system for identifying potential invaders than the conventional risk assessment methods.Item Augmented Reality to Reconstruct Sri Lankan Cultural Heritage in Prime State: HeladivaAR(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Galmangoda, G.D.D.M.; Gajanayake, P.Y.S.; Indika, K.P.; Rajapaksha, N.R.; Jayaweera, Y.Sri Lanka is one of the few countries in the world with a great history and cultural heritage which has been a spotlight in the tourism industry. These places attract many of the local as well as foreign visitors because of their historical value. They try to create an imaginary picture of how it could have been all those years back which could be a false image or a much less correct image than the actual image. Which is why this project is unique in the way it represents the ancient historical sites of Sri Lanka as they were at the time where they were in full structure. “HeladivaAR” is mobile phone application offers personalized augmented reality tours of archaeological sites. It uses image processing, 3D modeling, tracker identification using Android platform, historical books and views from historians and augmented reality techniques to enhance information presentation, reconstruct ruined sites, as it was on top of the existing ruins. By means of Augmented Reality, the real scene is enhanced by multimedia personalized interactive information to increase the experience of the user, who can retrieve this information by a user-friendly interface through their mobile phone. In education, virtual heritage becomes a platform for learning, motivating and understanding of certain events and historical elements for the students and researchers. This application provides a better understanding of Sri Lankan cultural heritage and lets users gain interactive knowledge on archeological facts of ancient kingdoms.Item Sensitivity of Acute Angle to the Heat Transfer Process of a Circular Tube with a Hemispherical Cup at the bottom of an Evacuated Solar Collector(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Kumari, K.D.N.; Wijerathna, J.K.The evacuated solar collectors are made by attaching parallel circular tubes with a hemispherical cup at the bottom and generally solar collectors are installed with an acute –angle to horizontal. Thus the present study is focused on the investigation of the sensitivity of the acute angle on heat transfer process in a circular tube in evacuated- solar collector. The heat and fluid flow is assumed to be unsteady, two-dimensional, laminar and incompressible. By giving uniform heat flux 950Wm-2at the upper lateral surface of the tube for a period of 15minutes, the velocity and temperature profiles of the water particles within the tube are studied for several acute angles /8, /6, /4, /3 and 3/8. For discretization of the governing equations, the finite volume method is used and as a numerical tool, the OpenFoam software, and the merged PISO - SIMPLE (PIMPLE) algorithm is used. The results show that, for small acute angles, although the maximum temperature gain is high, the majority of water particles are at low temperature level as well as the velocity profile is at a low level. For large acute angles, the velocity profile is high but maximum temperature gain is low. For moderate level acute angles, the maximum temperature gain is in somewhat high level and a lot of water particles are in high temperature level. Moreover, the velocity profile is in high level. Thus with a moderate level acute angle, the heat transfer process of water through a circular tube with a hemispherical cup at the bottom of the tube improves the performance.Item Natural Frequencies of Equilateral Triangular Plates under Classical Edge Supports(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Pradhan, K.K.; Chakraverty, S.In this paper, free vibration of thick equilateral triangular plates subject to classical boundary conditions has been investigated based on a new shear deformation theory, which needs no shear correction factor for stress distributions. The numerical modeling is performed by means of Rayleigh-Ritz method to obtain the concerned eigenvalue problem. The objective is to find the effect of different physical and geometric parameters on natural frequencies of the plate. New results for natural frequencies along with 3D mode shapes have been evaluated after the test of convergence and validation with the available results.Item Material Cost Estimation of Newspaper Printing Process by using Multiple Linear Regression Models(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Jayasinghe, J.M.T.N.; Abeygunawardana, R.A.B.Newspaper printing industry is one of the popular industry in developing countries like Sri Lanka. Newspapers become one of the major sources of disseminating news among people. When printing a newspaper, the company has to bear a huge cost since most of the newspapers are printed on a daily basis. Labour cost, material cost, machinery cost and electricity cost are major costs involved in newspaper printing cost. Among those costs, material cost is a combination of the various types of costs. According to the industrial investigations, material cost is a combination of the amount of paper, plate and ink usage in printing, number of printed newspapers and number of pages in a newspaper. The main objective of this research is to identify the contribution of the investigated factors on the material cost. Data were taken from a popular newspaper company in Sri Lanka for a popular daily newspaper. To analyse the weekly data for the material cost the data on the selected factors of; total number of pages, amount of paper, the amount of plate and ink, total printed newspapers were taken from the period of 06th January 2013 to 05th September 2014. Multiple linear regression models were fitted to identify the factors associated with the material cost. Among them, most suitable variables for the models were identified and the model was fitted. A number of pages per newspaper, the amount of printed newspapers, paper amount (kg) and ink amount (kg) becomes the significant factors associated with the material cost of newspaper printing. The amount of the variance explained by the model is about 78%. All the assumptions in multiple linear regression models were held in this model. Therefor we can use this model to estimate the material cost for a newspaper.Item A Model for Predicting Yield of Seeds in Ficus Fruits(Department of Statistics & Computer Science, University of Kelaniya, Sri Lanka, 2016) Jayarathna, H.L.D.K.; Nawarathna, L.S.; Karunaratne, W.A.I.P.Ficus is one of the largest plant genus which has an ecological significance due to the presence of “keystone” species. Availability of its sole mutualistic wasp pollinator and the effect of non-pollinator wasps determine the availability of seeds in Ficus fruits to produce the next generation of each species. In most of the previous studies on the yield of seeds of Ficus fruits, seeds have been counted manually, which is a time consuming and hectic process. Therefore, the main objective of this study is to introduce a model for predicting the yield of seeds in two Ficus species. Local polynomial regression, generalized additive models, and Poisson regression models were used for constructing these models to predict the number of seeds per fruit. Two generalized additive models which were constructed for Kandy municipal & Thumpane were best described with lower mean square error values of testing samples and moderately large R2 values when Fruit length was taken as a single predictor for both models. Poisson regression model gives a better result for modeling in Ficus callosa with min-max scaled variables, with a lowest mean square error value of the testing sample. Models which were built up for areas give the best prediction values when the yield of seeds less than 1000. By using local polynomial regression curves, it was identified that both biasedness and variance can be optimized using optimal bandwidth which was calculated by plug in the rule and it gives freedom to the flow of data by keeping non-parametric qualities.