Monthly Archives: November 2015

Panel Discussion on IoT-enabled Data Analytics

We would like to highlight a panel discussion on loT (Internet of Things) – enabled Data Analytics: Opportunities, Challenges and Applications at the INFORMS annual meeting that was held on Nov. 1st – Nov. 4th, in Philadelphia, PA.

In particular, four prestigious panelists joined this session, including Prof. Benoit Montreuil (Georgia Institute of Technology), Prof. George Q. Huang (Hong Kong University), Prof. Soundar Kumara (Penn State University), and Prof. Diego Klabjan (Northwestern University) who all have been the leading scholars in this field. The moderator of this session was Prof. Kaibo Liu (University of Wisconsin – Madison).

The goal of this session is to push the frontier in IoT application and the enabled data analytics research.… Read more

A Story of Basis and Kernel – Part II: Reproducing Kernel Hilbert Space

Author: Changyue Song

1. Opening Words

In the previous blog, the function basis was briefly discussed. We began with viewing a function as an infinite vector, and then defined the inner product of functions. Similar to \( \mathcal{R}^n \) space, we can also find orthogonal function basis for a function space.

This blog will move a step further discussing about kernel functions and reproducing kernel Hilbert space (RKHS). Kernel methods have been widely used in a variety of data analysis techniques. The motivation of kernel method arises in mapping a vector in \( \mathcal{R}^n \) space as another vector in a feature space.… Read more