moving average model

A moving average model attemtps to build a oredictor for one time series or {[sequwnce}} y_i as a moving average of another x_i. That is, it attempts to find weights w_k (0≤k<N) such that:

    y_i = w₀x_i + w₁x_i-1 + w₂x_i-2 + ... w_N-1x_i-(N-1) + ε_i

where ε_i is an error term that the model tries to minimise. Multi-linear regression can be used to find the weights w_k that minimise the sum of squares of ε_i.