Improving Performance of a Group of Classification Algorithms Using  Resampling and Feature Selection  Mehdi Naseriparsa, Amir-masoud Bidgoli, Touraj Varaee Islamic Azad University, Tehran North Branch, Department Of Computer  Engineering, Iran. Abstract— in recent years the importance of finding a meaningful pattern from huge  datasets has become more challenging. Data miners try to adopt innovative methods  to face this problem by applying feature selection methods. In this paper we propose  a new hybrid method in which we use a combination of resampling, filtering the  sample domain and wrapper subset evaluation method with genetic search to reduce  dimensions of Lung-Cancer dataset that we received from UCI Repository of Machine  Learning databases. Finally, we apply some well- known classification algorithms  (Naïve Bayes, Logistic, Multilayer Perceptron, Best First Decision Tree and JRIP) to the resulting dataset and compare the results and prediction rates before and after the  application of our feature selection method on that dataset. The results show a  substantial progress in the average performance of five classification algorithms  simultaneously and the classification error for these classifiers decreases  considerably. The experiments also show that this method outperforms other feature  selection methods with a lower cost.  Keywords-Feature Selection; Reliable Features; Lung-Cancer; Classification  Algorithms. Simulation of Improved Academic Achievement for a Mathematical  Topic Using Neural Networks Modeling  Saeed A. Al-Ghamdi, Hassan M. H. Mustafa Faculty of Engineering, Al-Baha University, Al-Baha, Kingdom of Saudi Arabia. Abdel Aziz M. Al-Bassiouni, Telecommunication & Technology Company, Cairo,  Egypt. Ayoub Al-Hamadi, Institute for Information and Communication Technology, Otto-  von-Guericke-University Magdeburg, Germany. Abstract— This paper is inspired by the simulation of Artificial Neural Networks (ANNs)  applied recently for evaluation of phonics methodology to teach the children “how to  read?” Nevertheless, in this paper, a novel approach is presented aiming to improve  the academic achievement in learning children as an adopted mathematical topic  namely long division problem. That's by comparative study of practical application  results at educational field (a children classroom); for two computer aided learning  (CAL) packages versus classical learning (case study). Presented study highly  recommends the novel application of interdisciplinary teaching trends as a measure  for learning performance evaluation. It is based on ANNs modeling, memory  association, behaviorism, and individual’s learning styles. Interestingly, observed and  obtained practical findings after the field application, proved the superiority of the  package associated with teacher's voice over both without voice, and classical  learning / teaching as well.   Keywords-Artificial Neural Networks; Learning Performance Evaluation; Computer  Aided Learning; Long Division Process; Associative Memory. A Conceptual Nigeria Stock Exchange Prediction: Implementation Using  Support Vector Machines-SMO Model  Abubakar S. Magaji, Faculty of Science, Kaduna State University, Nigeria. Victor Onomza Waziri, Audu Isah, Adeboye K.R. Federal University of Technology Minna-Nigeria.  Abstract— This paper is a continuation of our research work on the Nigerian Stock  Exchange (NSE) market uncertainties, In our first paper (Magaji et al, 2013) we  presented the Naive Bayes algorithm as a tool for predicting the Nigerian Stock  Exchange Market; subsequently we used the same  transformed data of the NSE and  explored the implementation of the Support Vector Machine algorithm on the WEKA  platform, and results obtained, made us to also conclude that the Support Vector  Machine-SOM  is another algorithm that provides an avenue for predicting the Nigerian Stock Exchange. Keywords- Nigerian Stock Market; Prediction; Data Mining; Machine Learning; Support  Vector Machine.