Modelling (Module A)
The first module, namely Modelling (Module A), focuses on the first phase of the MLOps lifecycle and is related to the lowest maturity level of the application of Machine Learning (ML) in organizations: modelling data. It includes the activities that form the basis of the application of ML, such as data extraction, data analysis, data preparation, model training and (mainly manual) model validation and evaluation.
In this phase, the focus is on correctly analyzing and modelling the data to achieve the business objectives and little use is made of automation (e.g. CI/CD), which is only added in the second phase of MLOps (Deployment – Module B). The modelling activities are often characterized by the manual, script-driven and interactive method by which the data analysis, preparation, model training and validation are carried out. To maintain an overview of the different models, parameters and choices that are being experimented with, experiment tracking is used.
From an ethical perspective, it is important in the modelling phase to devote sufficient time and attention to finding out the client’s objectives, mapping the stakeholders and exploring how the individual values of these stakeholders are affected (and recognizing possible conflicts between them). Aspects such as transparency, inclusion, security and privacy are of great importance in this. Naturally, attention must also be paid to the social and moral desirability of the client’s objectives. In addition, it is important to have (timely) awareness of possible biases/prejudices in the available data, recognise the possible consequences of these prejudices and find mitigations to deal with these prejudices.
Learning Outcomes
|
The student evaluates various ML techniques to make a well-founded choice, matching the required requirements of the custormer and implementing a prototype of the chosen ML technique to advise on solving a given data modelling problem
|
|
The student argues, using fundamental ethical frameworks how moral dilemmas can be solved and evaluates the possible consequences of existing biases in data and the influence of designed mitigations to counteract the consequences of this biases.
|
|
The student applies quantative and qualitative research methods to scientifically substantiate their choices during the ethical consideration(s) and making of the prototype.
|
| LEARNING OUTCOME 1 |
|
LEARNING OUTCOME 2 |
|
LEARNING OUTCOME 3 |
More information
Click here for an overview of all lesson plans of the master human centred AI
Please visit the home page of the consortium HCAIM
Acknowledgements
 |
The Human-Centered AI Masters programme was co-financed by the Connecting Europe Facility of the European Union Under Grant №CEF-TC-2020-1 Digital Skills 2020-EU-IA-0068.
The materials of this learning event are available under CC BY-NC-SA 4.0
|
|
The HCAIM consortium consists of three excellence centres, three SMEs and four Universities
|
