# Behavior Analysis through Routine Cluster Discovery in Ubiquitous Sensor Data

TL;DR: This work proposes a novel clustering technique for BA which can find hidden routines in ubiquitous data and also captures the pattern in the routines and efficiently works on high dimensional data for BA without performing any computationally expensive reduction operations.

Abstract: Behavioral analysis (BA) on ubiquitous sensor data is the task of finding the latent distribution of features for modeling user-specific characteristics. These characteristics, in turn, can be used for a number of tasks including resource management, power efficiency, and smart home applications. In recent years, the employment of topic models for BA has been found to successfully extract the dynamics of the sensed data. Topic modeling is popularly performed on text data for mining inherent topics. The task of finding the latent topics in textual data is done in an unsupervised manner. In this work we propose a novel clustering technique for BA which can find hidden routines in ubiquitous data and also captures the pattern in the routines. Our approach efficiently works on high dimensional data for BA without performing any computationally expensive reduction operations. We evaluate three different techniques namely LDA, the Non-negative Matrix Factorization (NMF) and the Probabilistic Latent Semantic Analysis (PLSA) for comparative study. We have analyzed the efficiency of the methods by using performance indices like perplexity and silhouette on three real-world ubiquitous sensor datasets namely, the Intel Lab Data, Kyoto Data, and MERL data. Through rigorous experiments, we achieve silhouette scores of 0.7049 over the Intel Lab dataset, 0.6547 over the Kyoto dataset and 0.8312 over the MERL dataset for clustering.

...read more

##### References

27,392 citations

### Additional excerpts

...such data, we first show a comparative analysis of three topic models namely Latent Dirichlet Analysis (LDA) [3] which is a generative statistical model where a set of observations are explained by unobserved groups....

[...]

25,546 citations

20,046 citations

2,812 citations

### "Behavior Analysis through Routine C..." refers methods in this paper

...So actually multiplying the three matrices gives P (w|d) = ∑ t Z P (w|t)P (t|d) Let number of documents be m and size of vocabulary be n, i.e A have a dimension of m × n, After SVD, dimension of U will be m× k, S will be k × k and of V will be k × n. 2) Non-Negative Matrix Factorization (NMF): This is also a matrix factorization method similar to SVD, with the constraint that the matrices have to be non-negative....

[...]

...SVD gives three matrices decomposed from the input as the output....

[...]

...For the implementation of PLSA, we use Singular Value Decomposition (SVD) [13] to factorize the matrix, allowing us to work with matrices of lower dimensions....

[...]

2,233 citations