Title：Real Applications of Optimization on Data Science
报告人：Wang Li （美国德克萨斯州立大学阿灵顿分校）
时间：2019.06.15上午9:00-9:40
地点：数学院203报告厅
Abstract: With the rapid development of science and technology, more and more datasets are produced for various purposes in the real world. For example, the cameras from the widely used mobile phones produce millions of photos daily. Mining these photos becomes very valuable for commercial and personal uses. High-throughput sequencing technologies produce tons of gene expression data. Analyzing these datasets is crucially important for improving public health and medicines. Often these datasets have hidden special structures embedded in their intrinsic low-dimensional spaces, while the observed data lives in high-dimensional spaces. The graph structure of data is generally unknown and difficult to obtain.
In this talk, I will introduce some optimization models that were proposed to build graph structures from data. Our goal is learning graph structures from data in the unsupervised learning, either in original high-dimensional space or reduced low-dimensional space. Some real applications will be discussed.
