CART, J-48graft, J48, ID3, Decision Stump and Random Forest: A comparative study
Taram Nayab Shah, Muhammad Zakir Khan, Mumtaz Ali, Bilal Khan, Nauman Idress
Abstract: Data mining techniques are very useful in the discovery of the hidden knowledge amid a huge amount of data, as will for merging similar data objects. Different algorithms and techniques are available for data mining like clustering, classification, association rule mining and neural networks to solve problems of data discovery and arrangement. The discussed algorithms are supervised learning whose labels are defined, in which classification is the most well-known method. Classification is a widely used method for a number of useful applications like artificial intelligence, credit card rating and fraud detection, etc. A number of Weka classifiers families are available such as Bayes, Lazy, functions, Meta, misc, rules and tree with their own pros and cons. Amid these algorithms, decision trees are the simplest and easiest algorithms for understanding, decision making and to compare with others due to hierarchal structure in nature. There is a number of Decision Tree algorithms used and employed by researchers that are available in the literature. However, this study focuses on the comparison of six decision tree algorithms that are CART, J-48graft, J48, ID3, Decision Stump and Random Forest. The objective of this study is to compare various decision tree algorithms to conclude the best algorithm for a particular problem using Python and Weka tool.
Fog Computing: A Bridge to the Cloud Computing in IoT Environment
Yasir Ali1
Abstract: Fog computing is also known as edge computing or Fogging. It is emerging technology and that provides an extension and enhancement to the cloud model of computing which attempts to bring the services, processing, application and storage to the edge of the network. This model helps in minimizing the latency and network congestion. Both technologies i.e. cloud and fog furnish similar resources and services, but the fogging is characterized by low latency with a wider spread and geographically distributed nodes to support quality and time period interaction. The basic purpose of fog computing or edge computing is to reduce the burden on cloud by bringing the workloads, services, storage facility, processing and bulky data close to network edge. In this paper, the tendency is to describe the supremacy of fog computing design over cloud computing technology in terms of architecture and other features.
The Impacts of Cognitive Elicitation Techniques in Requirement Elicitation Process
Yasir Ali1
Abstract: The most daunting task faced by RE is to understand the needs of stakeholders in more precise and complete ways. In RE, the end users desire for a system that may fulfill their needs but it is hard for them to describe these needs in more suitable and understandable manner due to the fact that the stakeholder’s viewpoints about system are rapidly changing as a result of rising evolution in requirements, cultural and political changes, various communication paths, linguistic barriers and some other factors. We need some methodology or meticulous requirement elicitation techniques that derive the actual need of end user from the system. This research paper is intended that how cognitive psychology and its elicitation techniques helps in describing the needs of various stakeholders and provides a sense of understandability for both parties i.e. stakeholders and requirement Engineer in the requirement elicitation process. The cognitive psychology digs into the mind of stakeholder by capturing all requirements that required to be replicated the proposed system by applying various cognitive techniques like protocol analysis, laddering, card sorting and repertory girds. In this paper, the main focus is to highlight the importance of cognitive elicitation techniques along with their advantages and disadvantages and its impact upon the RE process
OTES- Online Teacher Evaluation System
Saifullah, Asad Ali and Muhammad Bilal
Abstract: Online teacher evaluation system plays an important role in the university. At the end of every semester the Quality Enhancement Cell (QEC) of the university performs a teacher evaluation, but they are facing someproblems like security, time consuming and incorrect results evaluation as their system is manual and are not computerized. This project is a web based application through which the student can access it from everywhere when admin enable it for evaluation and save their time to fill the proforma efficiently. When students fill the teacher proforma and click the submit button then the record will save in database which is secured. In this website, we have two main parts; the one is the administration section which covers all the administrative data and entries made by students regarding teacher evaluation. The administrator can add/update the new departments, programs, sessions, teacher names and subject names to the website. The admin can also add specific programs with specific department and also specific subject give to specific teacher, one teacher can teach more than one subject. The second one is student section in which student can access the online teacher evaluation system and select their department along with program, session, semester, subject name and teacher name. After that, the students fill the teacher evaluation proforma including of 20 questions from different perspectives and then submit it. At the end, the record is saved in the database and automatically calculates the results and the admin can view and print the result. The proposed system can save resources like paper print, time and efforts.
