There is a large academic field focused on related work, but I will point to our work on modeling transformative AI risk. There is also work done more informally by groups of forecasters on Metaculus, for example, using shared google docs and collaborative meetings, and there is work on Delphi techniques for forecasting question development.
In addition, there is a large literature on eliciting Bayesian networks from groups of experts, often by using both data and expert input, and to perform similar tasks. I’m unsure what you are working on, but I’m happy to talk more about this and provide references and suggestions for who else to contact.
Thanks. I’m not working on anything at the moment, just curious about what has been done in the area. Did you consider other approaches to mapping out key hypotheses and cruxes for MTAIR? Do you have an idea of what advantages and disadvantages you expect the big Bayesian network to have compared to other approaches? Have you found it to be better or worse in any particular ways?
A particular question I’m curious about: have you found the big Bayesian network approach is helpful in terms of decomposing the problem into sub-problems and efficiently allocating effort to subproblems ?
We looked at the options, and choose Analytica largely because of a paucity of other good options, and my familiarity with it. Having spoken with them in the past, and then specifically for this project, I also think that the fact that the company which makes it is happy to be supportive of our work is a big potential advantage.
Why Not Large BNs?
BNs are expensive to elicit. (You need output∗∏inputi values elicited per node, where inputi and output are the number of discrete levels of each.) They also have relatively little flexibility due to needing those discrete buckets. There are clever ways around this, but they are complicated, and outside the specific expertise of our group.
BNs assume that every node is a single value, which may or may not make sense. Most software for BNs don’t have great ways to do visual communication of clusters, and AFAIK, none have a way to leave parts undefined until later. You also need to strongly assert conditional independence, and if that assumption is broken, you need to redo large parts of the network.
The way we actually did this shares most of the advantages in terms of decomposition and splitting subproblems, though removing duplication /overlap is still tricky.
Just to clarify: my understanding was that the MTAIR graph will eventually be extended with conditional probability estimates, so the whole model will define a probability distribution with conditional independences compatible with the underlying graph. This would make it a Bayesian Network in my eyes. However, it seems that we disagree on at least one of thing here!
Is the Analytica approach more robust to missing arrows not corresponding to conditional independences than a Bayesian network? If so, I’d be curious to hear a simplified explanation for why this is so.
Analytica allows you to define algebraic or other relationships between nodes, which can be real-valued, and have more complex relationships—but it can’t propagate evidence without explicit directional dependence. That allows more flexibility—the nodes don’t need to be conditionally independent, for example, and can be indexed to different viewpoints. This also means that it can’t easily be used for lots of things that we use BNs for, since the algorithms used are really limited.
There is a large academic field focused on related work, but I will point to our work on modeling transformative AI risk. There is also work done more informally by groups of forecasters on Metaculus, for example, using shared google docs and collaborative meetings, and there is work on Delphi techniques for forecasting question development.
In addition, there is a large literature on eliciting Bayesian networks from groups of experts, often by using both data and expert input, and to perform similar tasks. I’m unsure what you are working on, but I’m happy to talk more about this and provide references and suggestions for who else to contact.
Thanks. I’m not working on anything at the moment, just curious about what has been done in the area. Did you consider other approaches to mapping out key hypotheses and cruxes for MTAIR? Do you have an idea of what advantages and disadvantages you expect the big Bayesian network to have compared to other approaches? Have you found it to be better or worse in any particular ways?
A particular question I’m curious about: have you found the big Bayesian network approach is helpful in terms of decomposing the problem into sub-problems and efficiently allocating effort to subproblems ?
We looked at the options, and choose Analytica largely because of a paucity of other good options, and my familiarity with it. Having spoken with them in the past, and then specifically for this project, I also think that the fact that the company which makes it is happy to be supportive of our work is a big potential advantage.
Why Not Large BNs?
BNs are expensive to elicit. (You need output∗∏inputi values elicited per node, where inputi and output are the number of discrete levels of each.) They also have relatively little flexibility due to needing those discrete buckets. There are clever ways around this, but they are complicated, and outside the specific expertise of our group.
BNs assume that every node is a single value, which may or may not make sense. Most software for BNs don’t have great ways to do visual communication of clusters, and AFAIK, none have a way to leave parts undefined until later. You also need to strongly assert conditional independence, and if that assumption is broken, you need to redo large parts of the network.
The way we actually did this shares most of the advantages in terms of decomposition and splitting subproblems, though removing duplication /overlap is still tricky.
Just to clarify: my understanding was that the MTAIR graph will eventually be extended with conditional probability estimates, so the whole model will define a probability distribution with conditional independences compatible with the underlying graph. This would make it a Bayesian Network in my eyes. However, it seems that we disagree on at least one of thing here!
Is the Analytica approach more robust to missing arrows not corresponding to conditional independences than a Bayesian network? If so, I’d be curious to hear a simplified explanation for why this is so.
Analytica allows you to define algebraic or other relationships between nodes, which can be real-valued, and have more complex relationships—but it can’t propagate evidence without explicit directional dependence. That allows more flexibility—the nodes don’t need to be conditionally independent, for example, and can be indexed to different viewpoints. This also means that it can’t easily be used for lots of things that we use BNs for, since the algorithms used are really limited.