what is adaptive compressive sampling?

I have just started my work in compressive sensing. the measurement vectors are obtain by multiplying the sensing matrix with input signal. the thing i cant figure out adaptive compressive sampling.

can anyone please explain it as practical ?

thank you

In "ordinary" compressive sensing you take measurements $y = Ax$ and try to reconstruct $x$ from $y$ where $A$ - of dimension $(M \times N)$ - has been generated with no knowledge of x, except that it is sparse. $A$ is typically generated randomly.
In adaptive compressed sensing, you in principle take each individual measurement $y_m,\ m = 0, 1,\ldots M$, trying to adapt you next measurement $y_m$ cleverly so you get the best possible measurement for reconstruction of $x$, given what you have already observed in $y_0, y_1,\ldots, y_{m-1}$. Effectively, this means determining the rows $A_m$ of $A$ one at a time, taking into account the measurements that previous rows $A_0, A_1,\ldots, A_{m-1}$ gave you.