Is it possible to create an adaptive filter using neural networking toolbox so that after training it can filter noisy signal and give desired output? Been told it is called "denoising autoencoder"

Suppose my discrete time series containing noise signal is Yn(t) and my desired signal is Xn(t). I have 28 set of Yn(t)and corresponding Xn(t). Can I remove the noise signal using neural network, also one more problem I am having the time series data has only one feature (attributes). Is this solution possible using Neural Networking Toolbox whether I use use an filter or not?