Once a project-management use case has been selected see here, then the following steps can be applied to apply the approach to the projects.

  1. Selection of machine-learning method
  2. Selection of machine-learning model
  3. Selection of model environment
  4. Selection of machine-learning code library
  5. Selection of suitable data-sets for each Use Case
  6. Summary
  7. Get started
  8. Acknowledgements

Selection of machine-learning method

Use cases will be better suited to one type of ML method or another.

Selection of machine-learning model

To apply this type of ML method, one may wish to use one of the following specific model types as a first pass.

Type of ML model suitable per Project Use Case. Models that are classical rather than machine learning in brackets

Selection of model environment

Here are some accessible environments where one can apply these types of ML methods.

Selection of machine-learning code library

Here are some examples of libraries where these models can be found in a convenient format

Selection of suitable data-sets for each Use Case

For example, these data-sets could be selected.

Summary

Machine learning boosts most stages of the project lifecycle, from Project definition to close-out.

This has illustrated some routes to first application of machine learning models with one’s own portfolio data.

  1. Identify possible use cases
  2. Map to business area/lifecycle
  3. Narrow down to a generic ML method
  4. Select a straightforward ML model
  5. Choose a suitable model environment
  6. Choose a code library or algorithm to apply that model
  7. Select data-set

This gives something to play with, and get something to work. There are plenty of ML prototyping frameworks available to provide structure for developing a prototype.

Get started

Please go to the relevant code and document libraryhere

Acknowledgements

Agrawal et al </a> have a canvas here for planning it at high level, understanding purpose and constraints.