Welfare decisions being made by algorithms 13:16 - Oct 15 with 839 views | StokieBlue | I am quite surprised that so many councils are using machine learning to make some of these decisions. There are a number of issues around overfitting which I would think would be magnified given the likely patchy data that councils have. https://www.theguardian.com/society/2019/oct/15/councils-using-algorithms-make-w "But concerns have been raised about privacy and data security, the ability of council officials to understand how some of the systems work, and the difficulty for citizens in challenging automated decisions." “There is too much hype and mystery surrounding machine learning and algorithms. I feel that councils should demand trustworthy and transparent explanations of how any system works, why it comes to specific conclusions about individuals, whether it is fair, and whether it will actually help in practice.” This is especially important as it's not always possible to give a transparent explanation of how the algorithm reached it's decision and thus feeds back into the point of it being hard to challenge. It just all seems a bit premature in a rush to cut costs and use the latest tech. SB | |
| Avatar - IC410 - Tadpoles Nebula |
| | |
Welfare decisions being made by algorithms on 13:22 - Oct 15 with 822 views | DanTheMan | Such a young field, weird seeing it already being used this way and for thing seemingly so important. You're of course right about the explanations. That's not how machine learning works, it's all essentially based on models, not like it's going to print out a list of reasons why it's predicted something. It just has based on the training data. And if your training data is wrong... well. | |
| |
Welfare decisions being made by algorithms on 13:48 - Oct 15 with 777 views | flimflam |
Welfare decisions being made by algorithms on 13:22 - Oct 15 by DanTheMan | Such a young field, weird seeing it already being used this way and for thing seemingly so important. You're of course right about the explanations. That's not how machine learning works, it's all essentially based on models, not like it's going to print out a list of reasons why it's predicted something. It just has based on the training data. And if your training data is wrong... well. |
Written in Prolog | |
| All men and women are created, by the, you know the, you know the thing. |
| |
| |