Big Data is the Buzzword of 21st century as we know it and has been extremely useful in several risk assessment tasks. The effectiveness of Big data on risk management depends on accuracy,consistency ,completeness and timeliness of data. Some most common mistakes made by Big Data experts who are involved in risk management are : Confirmation Bias : It occurs when data scientists use limited data to prove their hypothesis.
Selection Bias : When data is selected subjectively, Analyst comes up with the questions and thus almost picking the data that is going to be received ( Ex : Surveys)
Outliers : Outliers are often interpreted as normal data
Simpson’s Paradox : When group of data points to one trend, but can reverse when they are combined
Confounding Variables are overlooked
Analyst assume bell curve
Overfitting and Underfitting models
Read more at : http://dataconomy.com/2017/01/7-mistakes-big-data-analysis/