Jauhar Ali

In this paper, we implemented an approach to predict final exam scores from early course assessments of the students during the semester. We used a linear regression model to check which part of the evaluation of the course assessment affects final exam score the most. In addition, we explained the origins of data mining and data mining in education. After preprocessing and preparing data for the task in hand, we implemented the linear regression model. The results of our work show that quizzes are most accurate predictors of final exam scores compared to other kinds of assessments.

The aim of this research paper is to recognize patterns and relationships in the data collected from World Bank Statistics using the “Correlation” data mining technique and the “Correlation Matrix”. The data used spans from the Year 2000-2005 and includes numerous countries of the world. It comprises of data related to the international economic and health statistics. The software Rapid Miner is used in order to analyze and evaluate the data. This paper aims to discover underlying patterns related to various attributes that are difficult to detect or predict.

In our day to day life, predictability of gold's price is significant in many domains such as economics, trading, investment, and financial and political environments. Better investment decision could be made when gold price values are accurately predicted. The main objective of this research is to forecasts whether the price of gold will rise or decline in the near future. Decision Tree, Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Linear Regression are all different classification algorithms that have been used in this paper to predict the gold price movement direction. The performance of each of these algorithms has been investigated while using the Rapidminer software.

We consider the problem of predicting direction of movement of popular currencies in the foreign exchange (Forex) market. Technical analysis is an analysis methodology for forecasting the direction of prices through the study of past market data. A currency strength indicator is a metric showing the relative strength of currencies over a period of time. In this study, we analyzed the performance of a strength indicator in a different way. Based on the change in strength, we opened a basket of virtual trades and measured their profitability. Several tests have been performed and they show consistent profits. The results are encouraging and a profitable forex trading system can be made by only considering strength changes in currencies.

An Industrial Case Study for Scrum Adoption

Statechart, which is one of the important diagrams in Unified Modeling Language (UML), represents reactive behavior of a class of objects. Implementing statecharts has been difficult for programmers because the commonly used object-oriented programming languages do not support statecharts at the language level. In this study, we present a new approach to implement statecharts using Java enums. Each state is represented as an enum-value. The whole statechart behavior is encapsulated within the Statechart class. This class is nested inside the owner class of the statechart. The resulting code is object-oriented, efficient and well-structured.

Data structures is a core course in almost all computer science programs. Textbooks and courses on data structures these days are programming-languages-based with those based on object-oriented programming (OOP) in dominance. In many cases, they base their pedagogy on software design patterns. However, because of heavy reliance on abstraction, they place greater demands on learners. In this study, we design and implement a visualization tool that makes it easier for students to understand object-oriented data structures.