Intelligently transforms the data for predictive modelling while retaining essential information in the data. Depending on the data and the specific block this includes: checking the relevance of the data with break detection and data reduction, transforming unstructured data into structured data, and selection of training, validation and test sets
A block may contain multiple data science or machine learning algorithms applicable for its specific goal. The model selector determines which algorithm(s) is the best match and gives the most accurate predictions for the specific business problem given the data currently available.
To ensure accuracy and reliability of the algorithm and outcomes, all predictions will be validated and checked before execution. When validated successfully, the predictive model will be used to extrapolate to new scenarios.