exercise set 6 question 3.3

exercise set 6 question 3.3

par Kiarash Farivar,
Number of replies: 3

I think the answer for this question in the solution is wrong .

is the answer the same as calculations in the last video of this week  (I think around 5:20)?

In reply to Kiarash Farivar

Re: exercise set 6 question 3.3

par Georgios Iatropoulos,

I can't see any errors in the solution. Could you be more specific about what you think might be wrong? As you point out, the calculation is exactly the same as in the lecture "2.5: Towards biology: Spiking neurons" (starting at around 5:00). Note that the sum over mu is dropped in the lecture (this is mentioned at 5:40). This sum is included, however, in the solution to exercise 3.3.

In reply to Georgios Iatropoulos

Re: exercise set 6 question 3.3

par Kiarash Farivar,

well the question mentions that we have low-activity patterns and gives the w_ij formula that involves ksi ,a and b but the solution replaces the w_ij with p_i and p_j.

Also the final answer in the video has grouped it's sigma representation of neurons into two sums one positive the other one negative this is not done in the solution.


In reply to Kiarash Farivar

Re: exercise set 6 question 3.3

par Georgios Iatropoulos,

I see now that you are looking at the solution in the courseware. This solution is indeed incorrect and outdated. We will update it. However, as Prof. Gerstner pointed out in an earlier post in this forum, you should stick to the course material in Moodle; this is the main website for the course. When we post the solution sheet to exercise set 6, it will contain the correct solution to question 3.3.