1: \begin{abstract}	

2: 	Structural Health Monitoring (SHM) provides an economic approach

3: 	which aims to enhance understanding the behavior of structures

4: 	by continuously collects data through multiple networked

5: 	sensors attached to the structure. This data is then utilized to gain

6: 	insight into the health of a structure and make timely and economic

7: 	decisions about its maintenance. The generated SHM sensing data

8: 	is non-stationary and exists in a correlated multi-way form which

9: 	makes the batch/off-line learning and standard two-way matrix

10: 	analysis unable to capture all of these correlations and relationships.

11: 	In this sense, the online tensor data analysis has become an essential

12: 	tool for capturing underlying structures in higher-order datasets

13: 	stored in a tensor $\mathcal{X}  \in  \mathbb{R} ^{I_1 \times \dots   \times I_N} $. The CANDECOMP/PARAFAC

14: 	(CP) decomposition has been extensively studied and applied to

15: 	approximate X by N loading matrices A(1), . . . ,A(N) where N represents

16: 	the order of the tensor. 	We propose a novel algorithm, FP-CPD, to parallelize  the CANDECOMP/PARAFAC (CP)  decomposition  of a tensor  $\mathcal{X}  \in  \mathbb{R} ^{I_1 \times \dots   \times I_N} $.  Our approach is based  on stochastic gradient descent (SGD) algorithm  which   allows us to parallelize the learning process and it is very useful in  online setting since it  updates $\mathcal{X}^{t+1}$ in one single step. Our SGD algorithm is augmented with Nesterov's Accelerated Gradient (NAG) and perturbation methods to accelerate and guarantee convergence. The experimental results  using laboratory-based and real-life structural datasets indicate fast convergence and good scalability.

17: 		

18: 

19: \end{abstract}