[NETVIBES] Data Scientist

Functional Value

Machine Learning Workbench to build & publish Machine Learning Models and Procedures.

- Selection of pre-defined algorithms (classification method),

- Training Models with tabular data sets ingested in batch mode,

- Evaluating Models effectiveness looking at generated key metrics (confusion matrix, accuracy, precision, recall, ROC, ...),

- Testing several Models with different settings and comparing versions to find the optimum one,

- Publishing validated Models and Procedures to be made available for usage in production.

User Experience Value

Applying scientific methods, processes & algorithms to extract knowledge and insights from massive amount of structured data

I
Data Scientist application: - Machine Learning Workbench application to define, train and validate AI procedures (e.g. machine learning models) for production usage in Data Perspectives & applications.

I
Building New Models configuration step: - Dataset selection and previsualization.

I
Building New Models modeling task definition step: - Target definition and features selection.

I
Building New Models choosing learning algorithms step: - Selection and configuration of Machine Learning Models to be trained.