Browsing by Subject "Data mining"
Results Per Page
Sort Options
-
ItemClustering approach to examination scheduling( 2010-05) To, Tang Van ; Win, Soe SanThe difficulty in examination scheduling is to draw an examination timetable by taking a number of different constraints into account. This paper attempts to optimize two major constraints-to minimize the examination time conflicts and the number of consecutive examinations for a student in a day. Clustering method is applied by analyzing exam enrollments and then groups students into clusters so that students within a cluster are having more similar subject combination to each others than those in another clusters. The purpose of clustering is to effectively arrange the order of exams to be allocated into a suitable exam period. Since exams are scheduled by clusters, it solves the exam conflicts for the students within the same cluster whose are closely related to each other.
-
Item
-
Item
-
Item
-
ItemFrom Bits and Bytes to Big Data - An Historical OverviewAbout a century ago the computer revolution started in the USA. Computers and mobile devices have now become an integral part of our lives in the workplace. The first attempts to investigate the data output started in the 1960s on a command prompt blinking on the computer monitor. Later in the 1990s, powerful data mining tools were created to sort and analyze large amounts of data. Now the new technologies and tools such as Big Data Analytics allow us to instantly capture and analyze an enormous amount of personal and institutional data about clients, financial transactions, social networks, and everything else happening online. This paper attempts to provide a fresh insight into the evolution of “data” acquiring such a prominent role and value. It provides a pictorial history using secondary data sources and attempts to provide an overview of current tools, technologies, and trends associated with the evolution of Big Data Analytics.
-
Item
-
ItemA Self-Growing and Self-Organizing Batch Map with Automatic Stopping Condition( 2013) Kim, Se Won ; To, Tang VanThis paper proposes a model of self-growing and self-organizing feature map designed to alleviate the difficulty of predetermining an appropriate size and shape of the feature map suitable for the input data in the applications of the Self-Organizing Map. The proposed model progressively builds a feature map by incremental growing of the network in a way that maintains two-dimensional regular grid structure and gradual adaptation of the reference vectors by coordinated competitive learning dynamics of the Batch Map algorithm. Experimental results based on iris data set and Italian olive oil data set show that the proposed model is effective in discovering an appropriate size and shape of the network grid to manifest a suitable feature map for the input data and that the resultant feature maps are comparable to feature maps produced by the standard SOM algorithm in their quality.
-
Item
-
Item