10 2. a) For the case of single-layer perceptron (SLP) neural networks with linear activation functions: i) Analyse the linear least squares (LLS) training algorithm, when all input patterns and desired outputs are available at once. ii) Explain how the least mean squares (LMS) algorithm for training SLP networks when the patterns are not available at once, differs from the 6 LLS method. 10 2. a) For the case of single-layer perceptron (SLP) neural networks with linear activation functions: i) Analyse the linear least squares (LLS) training algorithm, when all input patterns and desired outputs are available at once. ii) Explain how the least mean squares (LMS) algorithm for training SLP networks when the patterns are not available at once, differs from the 6 LLS method