In Recitation 8 Ex1 we are asked to compute $v_{t}^{\lambda}$.

which is defined: $v_{t}^{\lambda} = (1 - \lambda)\sum_{n=1}^{\infty} \lambda^{n-1} v_{t}^{(n)}$

In class we defined that for n larger then the episode length we pad with zeros.

However in the recitation wev'e simply set $v_{t}^{(n)}$ to 0 for all $n$ larger then the episode.

The 2 approaches give different solutions.

(1 - Recitation)

$v_{1}^{\lambda} = 0.5\cdot[1\cdot(-1)+0.5\cdot0+0.25\cdot0.25]=-\frac{15}{32}$

(2 - considering all n up to infinity)

$v_{1}^{\lambda} = (1-\lambda)\sum_{n=1}^{\infty}\lambda^{n-1}v_{1}^{(n)} = (1-\lambda) [-1+0.5\cdot0+\sum_{n=3}^{\infty}\lambda^{n-1}0.25] = -\frac{7}{16}$

which is correct?

Thanks