(2/7): We propose a new approach to account for the fact that neither experimenters nor decision makers have direct access to “true” subjective values; only observed value estimates and choices.
Show this thread
Checkout our new article in @NatureNeuro : “Efficient coding of subjective value”.
https://www.nature.com/articles/s41593-018-0292-0 …
http://em.rdcu.be/wf/click?upn=lMZy1lernSJ7apc5DgYM8WIknn5c6YFKEPXVAUGxHSE-3D_3GEjPxJ3WsP8IaZacR37iDde7GkE-2Fq9geo3zS-2FVnK52ifGClaAOZYhHTzX8PAqjVIk7kJcxQLOcwH3EGxzPWZPRaKXSQkPVSgA5VFR6IysPcCx7Ggw5q-2FhgDJkzzYzwhKKyLyLdhRanB-2Bg4sFYbTGi3wS8YBgGzPwGTfPC9olIzwqN6S4GH0jUyqL2BQrj-2FKnD3xEw-2F773SRVbCDnaF4H5-2BOCq7HB0blQf5A23YI-2FnZyBxnbdHtp6lCTgXVuMUQEmS45D-2BQvJJTR6QXVXOGqqg-3D-3D …
Mini-Thread of key take-home messages (1/7).
@jgvfwstone
-
-
-
(3/7): This framework may serve to improve the modelling and prediction of any cognitive process that involves fully subjective value estimations (such as pain and health perception, to mention only two examples).
Show this thread -
(4/7): Accounting for a system with limited capacity, our model accounts for variability in value estimations, binary choices, and for subjective value estimation biases as a function of time pressure … and this extends to confidence ratings.
Show this thread -
(5/7):Our procedure differs from traditional economic approaches that derive preferences directly from observed choices, ignoring processes involved in estimating subjective values (and the associated sources of variability). Our results show that this ignorance is not warranted!
Show this thread -
(7/7): More generally, the results support our emphasis on the desirability of developing models of decision making that account simultaneously for the goals of the organism, its environment, and its biological constraints.
Show this thread -
IMPORTANT: Open PhD position to work on these topics in my new lab at ETH in beautiful Zurich. Preferentially students with strong background in computational neuroscience, statistical inference, etc. Write me if you are interested (http://decision.ethz.ch/ ). Please RT!
Show this thread
End of conversation
New conversation -
-
-
If it was Open Access I would read it!
-
Here is the shareable link https://rdcu.be/bdRix But if it does not work fine for me, just drop me an email and I am happy to share the article with you
End of conversation
New conversation -
-
-
Nice!’doubt the 2nd line of the abstract was meant to be political but it’s hard not to see it that way. Anyway, congrats!
-
Under the economists' hat mmm.... not too political ;)
End of conversation
New conversation -
-
-
Neat! Your efficiency is defined as bits/r; r=response. If r is bounded then mutual info is maximised by max entropy uniform p(r); I think you assume r is bounded. But if mean r is fixed => maxent p(r)=exp(-r) etc. Tested? Efficiency as bits/J, relevant? Twiter<ideal for nuance.
-
We are currently working on a related, but slightly different approach where the mean activity is fixed and where efficiency could be linked to metabolic costs. I will let you know as soon as we have this part ready. BTW, your books have always been an inspiration!
-
That sounds very fine. Looking forward to seeing your new work. And thank you. I have written only one book, but many times, I think.
End of conversation
New conversation -
-
-
Congrats Rafa! Awesome work and excellent mini-thread!
Interested in doing a PhD? @RafaPolania is hiring - great guy, great lab, great city, best department@ETH_en
#computational#neuroscienceThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.