statistics

Modern Techniques of Statistical Optimization for Machine Learning

Tong Zhang, Professor of Computer Science and Mathematics, HKUST

-

B9 H1

Many problems in machine learning rely on statistics and optimization. To solve these problems, new techniques are needed. I will show some of these new techniques through selected machine learning problems I have recently worked on, such as nonconvex stochastic optimization, distributed training, adversarial attack, and generative models.

Finding the Face in the Crowd

1 min read ·
An algorithm that maximizes the difference between data categories achieves high accuracy in classifying faces.

Pulling rank on spatial statistics

1 min read ·
A technique that uses the power of computing could solve statistical problems cheaper and faster than current methods.

Cutting datasets down to size

1 min read ·
A powerful statistical tool could significantly reduce the burden of analyzing very large datasets.

Training agents to walk with purpose

1 min read ·
By training a search agent to make smarter exploratory decisions, relational data can be classified more accurately and efficiently.

A model for millions of locations

1 min read ·
CEMSE statisticians developed a framework which enables modeling of a range of meteorological and environmental datasets from up to 2 million locations globally.

Abdulhakim Qahtan

Abdulhakim Qahtan obtained his Ph.D. degree in Computer Science under the supervision of Professor Xiangliang Zhang at Machine Intelligence and kNowledge Engineering (MINE) Laboratory at King Abdullah University of Science and Technology (KAUST). Dissertation Title: Efficient Estimation of Dynamic Density Function with Applications in Streaming Data. First employment: Postdoctoral researcher at QCRI in Qatar. Research Interests Abdulhakim's research interests included Data mining, Statistics​​​ and Applying statistical models in machine learning. Selected Publications Basma Alharbi, Abdulhakim