TASG: An Augmented Classification Method for Impersonal HAR

Algorithm 1

Activity sparse grouping algorithm.

Input: A: {A1, A2,…, An}, G = Ø

Output: G: {G1, G2,…, Gm}(1)Sparse decomposition of training data and solve the sparse coefficients;   (2)Calculating the similarity matrix ;   (3)for i = 1, 2, …, n      if then           end if  end for(4)Do:  for each       if Aq ∈ Gkthen    add Ap into Gk and remove Ap from A           end if  end for each until A = Ø or A no longer changes(5)if A ≠ Ø  cluster remaining activities in A into a new group Gm and add Gm into G     end if(6)Return G;