Perspectives and Conclusion

Tensor and tensor decomposition are very versatile and powerful tools, which has a wide range of applications in data mining, signal processing, computer vision and other fields [1–5]. For a third-order HDI tensor modeling a dynamic network, this book carry out some preliminary research on latent factorization of tensors methods to implement accurate representation for dynamic networks. Further, in real industrial applications, in order to tackle a variety of complex network analysis tasks, it is fundamental and necessary to build accurate representation of a complex network based on latent factorization of tensors via fusing advanced algorithms or introducing property information [6–10]. In particular, there are several research directions that can be continuously researched in the further: