W2D4 content discussion

[rosana]

In Tutorial 2, we got confused by the last statement in this image. It seems to say that a_t = a_bar_t + L_t * (s_t - g_t). Is this correct?

In the faculty Q&A, Dr. Saxena said that the controller action a_t = L_t * s_t and that it does not depend on a_bar_t.

What are the correct equations here?

[ido.benartzi]

Well it’s not really about the content, but I was curious about the " *\ " operator that was used in the “belief_update” function in tutorial1 and couldn’t find anything about it.

[ZaneMitrevica]

@ido.benartzi \ is used merely to split the statement into multiple lines so that the code easier to read. It does not affect the output, making * just regular multiplication.

[TQi]

Hi, I have the same question…is there an answer for that?

@ollimuh

I’m not sure I understand the change to the cost function. I would assume that the action cost (eg fuel expenditure) would depend solely on the final action a_t. Why should it ‘care’ about \bar a_t?

And I’m also not sure how to interpret the sum of \bar a_t and L_t(s_t-g_t). If the former is the action to get to g_{t+1} and the latter is the action to get to g_t, then linearly wouldn’t this sum let us get to g_t+g_{t+1}?
Thanks!