Thinking Statistically -- from small data to big data
摘要: Technological advances in recent years have resulted in a very fast accumulation of a huge amount of high-dimensional data in many fields, e.g., biomedical researches, finance, internets, e-commerce, etc, and give rise to the buzz word "Big Data" popular throughout the world. Recent successes of deep learning and other machine learning tools have also fueled the recent hype in artificial intelligence (AI). I believe that, besides their traditional roles in survey sampling and population studies, statistical concepts and methodologies are still key to understanding these AI tools, to realizing the full potential of the modern-day data, and to advancing future developments of new data models and AI tools. I will describe a few examples, mostly from my own research work, illustrating important roles of statistics in elevating these fields. They include computational biology, medicine, engineering, and economics.