To the content To the search functionality

DataOps Specialization

  • Code

    L.29631

  • Amount of hours required

    420

  • Language

    en-GB

  • Quartile of execution

    3

With the rise in possibilities for data analytics, automation is an increasingly important topic to create efficient and effective workflows. The concept of DataOps (Data and Operations) blends ideas from DevOps and agile development to create an automated, process-oriented methodology to improve the quality and reduce the cycle time of data analytics. It calls for both a technical and a business understanding of data-related challenges, specifically for the integration of data warehouses, processing, analytics, and evaluation. Key in this specialization is a full understanding of all facets of the DataOps paradigm and being able to implement its technical steps to effect automation of data analytics. Therefore, this specialization has two aims:

  • Students attain an understanding of the theory and background of statistics and linear algebra for data exploration and applied machine learning.
  • Students become fluent in the use of tools and frameworks for automation of the entire DataOps workflow, from data pre-processing given a data warehouse to the evaluation of results, as well as deployment.

Competences

  • Master ICT Software Engineering 11 SW/ANA / 4
  • Master ICT Software Engineering 12 SW/ADV / 4
  • Master ICT Software Engineering 13 SW/ONT / 4
  • Master ICT Software Engineering 14 SW/REA / 4
  • Master ICT Software Engineering 15 SW/M&C / 4
  • Master ICT Software Engineering Research Skills
  • Master ICT Software Engineering Prof. Skills

Learning goals

  • Understand the data analytics workflow (based on CRISP-DM (Wirth, & Hipp, 2000)).
  • Apply basic concepts from statistics to describe data (distributions, visualizations, and exploration).
  • Execute all steps in the data analytics workflow, namely: data understanding, data pre-processing, modelling, evaluation, and deployment.
  • Assess the suitability of different data modelling methods/algorithms for optimal performance and evaluate training results objectively.
  • Assess presence/absence of common pitfalls in data analytics in a use case (non-representative data, low-quality data, discrimination, overfitting, irrelevant features).
  • Evaluate the data analytic models by designing feedback possibilities to communicate model results.
  • Build an automated workflow for data analytics to ensure deployability and scalability.

Tests

Code Name
WC Werkcollege
T.55354 DataOps Specialization