Evaluate risk of deploying algorithmic models
About
This unit is about evaluating the risks in deploying algorithmic models and developing mitigation measures for internal and external clients.
Scope
define model scope, analyse algorithmic model
Define model scope
- define the purpose and metrics for the algorithmic model
- define data sources used to design the model and data flows used to reinforce the model
- define and evaluate the assumptions used while designing the algorithmic model
- evaluate the range of expected outcomes of the algorithmic model
Analyse algorithmic model
- test the model with different inputs and identify the factors that are creating a deviation from the expected outcomes
- estimate the risks involved in case the algorithmic model deviates from the expected outcomes
- introduce checks and mitigation measures for each of the potential risks resulting from the model
- create documentation on potential risks and the associated mitigation measures
- validate risks and mitigation measures with appropriate stakeholders
- recommend and implement corrective actions to the model as required
- evaluate the model for all possible use cases / scenarios
Required Knowledge & Understanding
Technical Skills
- different use cases and the suitability of various algorithmic models to address them
- different factors that contribute to risks. For example, outdated, irrelevant or biased input data, insufficient sample sizes, flawed design logic, coding errors etc.
- different mitigation measures along areas such as data selection, algorithm design, live use in production etc.
- how to identify and refer anomalies in data
- how to work on various operating systems such as linux, ubuntu, or windows
Soft Skills
Decision Making
follow rule-based decision making processes. make decisions on suitable courses
Analytical Thinking
analyze data and metrics for various use-cases and environments. impact analysis of the various actions performed and disseminating relevant information to others
Attention to Detail
check your work is complete and free from errors