AI – Maintenance and Update of the Dataset

This question is related to the “maintenance and update”. If these two characteristics are not guaranteed, the risk is to deal with information that isn’t valuable and thus usable, eventually. See also (Salay et al., 2017).

Main Question

Is the dataset considered to be maintened and updated in time, properly?

Sub-Questions

1. Can the dataset be updated with new data (in terms of time and space)?
2. Are potential inconsistencies resolved when merging data from different sources?
3. Is it often necessary to update the data to keep (or improve) the performance of algorithm?
4. Can the dataset be updated with new data? (e.g., extending the number of situations/conditions covered)?

References

• Salay, R., Queiroz, R. and Czarnecki, K. (2017) ‘An Analysis of ISO 26262: Using Machine Learning Safely’. Automotive Software. doi:https://doi.org/10.48550/arXiv.1709.02435.