|
Data and its accuracy are highly crucial for the success of a statistical data modeling process. It is highly unlikely that data modelling can make error-ridden data accurate, irrespective of how good the data modelling process or the professionals who do this are.
That is why we are involved with our clients right from designing the survey instrument to delivery and collection of data, which we then extract and analyze to come up with actionable recommendations.
Some of the techniques we use in statistical data modeling include regression analysis including logistic regression and survival analysis, multivariate techniques such as discriminate analysis, K-means clustering, and non-parametric methods.
We also use Estimation and hypothesis testing, Frequency / cross-tab / percentage analysis and Forecasting models in addition to Time series & trend data and ARIMA, exponential smoothing.
In addition, we also indulge in continuous monitoring of the performance of the models over time and develop new models when there has been degradation of performance. |