Stages

Data improvement

Data collection and quality control.

Data from the real world or from opinion surveys.

Datasets that will be enriched as the analysis progresses.

This stage requires a lot of time.

Example:

Basic Statistical Descriptions

Properties of the data.

Pivot tables.

Business Intelligence

Identification of atypical values: outliers, fraud… or innovative profiles based on new topics to be soon incorporated as habitual behavior (business opportunities).

Prediction

Regression-type methodologies. Classifiers.

Linear & nonlinear approaches.

Hypothesis testing.

Machine learning.

No blackbox: no unknown formula.

Implementation

Algorithm in production.

On premises or in the cloud.

Involved stakeholders.

Outside intervention.