International Journal of Latest Trends in Engineering and Technology (IJLTET)
The paper addresses some theoretical and sensible aspects of information mining, specializing in prophetical data processing, wherever 2 central forms of prediction issues are discussed: classification and regression. Additional accent is formed on prophetical data processing, wherever the time-stamped information greatly increases the size and complexness of downside resolution. The most goal is thru process of information (records from the past) to explain the underlying dynamics of the complicated systems and predict its future. Traditional classification algorithms may be restricted in their performance on extremely unbalanced datasets.