Dear Daniel,
first of all: Great book! I do enjoy reading it!
In Figure 8.6 for the input x and the hidden-input y both vectors are transformed with the weight matrix W. It looks like both are sharing the same weight matrix. Maybe it would be better to use Wi and Wh instead, to make clear that these are different weight matrices.
Best
Benjamin
Dear Daniel,
first of all: Great book! I do enjoy reading it!
In Figure 8.6 for the input x and the hidden-input y both vectors are transformed with the weight matrix W. It looks like both are sharing the same weight matrix. Maybe it would be better to use Wi and Wh instead, to make clear that these are different weight matrices.
Best
Benjamin