Fraud is one of the major issues in today’s business world. It was found that 8% increase over last year in the cost per dollar of fraud losses and online retailers are the worst hit as fraudulent transactions grew 32.1% in 2015. Nowadays, data analysis has come to rescue organizations from fraud as it allows companies to create fraud risk profiles and then use existing data to identify potential fraudulent activity. This article explores the commonly used data analysis techniques used to search out fraudulent transactions. They are: Pattern Recognition, Outlier Detection, Regression Analysis, Semantic Modeling, Neural Network, and Anomaly Detection. Read more at: http://it.toolbox.com/blogs/itmanagement/how-to-detect-fraud-using-data-analysis-74726

An investment banking firm  has developed a ‘stochastic model’ its predictions to generate distribution of results for the each of the 64 matches to be played during WC.The predictive model is part of the the World Cup and Economics 2014. Stochastic modeling is used to forecast outcomes, even in the field of sports. This mathematical model predicts the probability of outcomes. According to the model, predicting the outcome of the WC is based on predictions for each match on “regression analysis”. This analysis used 14,000 observations to approximate the coefficients of its model. The dependent variable is the number of goals attained by each team during a match. Based on this analysis Brazil fares as the winning team. The next are Argentina and Germany, but these two teams have inferior probability. Read more at:

http://blog.arekibo.com/world-cup-2014-the-use-of-analytics-to-predict-the-winning-team/