Computationally Efficient Methods for MDL-Optimal Density Estimation and Data Clustering

The Minimum Description Length (MDL) principle is a general, well-founded theoretical formalization of statistical modeling. The most important notion of MDL is the stochastic complexity, which can be interpreted as the shortest description length of a given sample of data relative to a model class. The exact definition of the stochastic complexity has gone through several evolutionary steps. The latest instantation is based on the so-called Normalized Maximum Likelihood (NML) distribution which has been shown to possess several important theoretical properties. However, the applications of this modern version of the MDL have been quite rare because of computational complexity problems, i.e., for discrete data, the definition of NML involves an exponential sum, and in the case of continuous data, a multi-dimensional integral usually infeasible to evaluate or even approximate accurately. In this doctoral dissertation, we present mathematical techniques for computing NML efficiently for some model families involving discrete data. We also show how these techniques can be used to apply MDL in two practical applications: histogram density estimation and clustering of multi-dimensional data.

Tsoi Yuen Resistance and Post-80s Generation: A new wave of domestic social movement

Hong Kong was once a British colony and has been under the sovereignty of People’s Republic of China (PRC) since 1997. However, some of the unjust practices and colonial legacies are infiltrated into the development ideology as well as the social structures. The construction of intercity express railway project announced in 2008 causing the demolishment of Tsoi Yuen Tsuen, a “non-indigenous” agricultural village in Hong Kong, was one of the current examples. Tsoi Yuen village was established under the former colonial sovereignty sixty years ago. Approximately 450 populations were affected that they had to relocate their homeland involuntarily. However, these villagers were very attached to their homelands and were unwilling to move, and meanwhile they found that they were absent in the government’s consultation and decision-making process. Soon they began their resistance and demanded for “No Move! No Demolish!”. Their movement was strongly supported by a group of “Post-80s generation” and turned into the most important social movement of the city in recent years. In fact, demolition of Tsoi Yuen Village for city development is not an isolated case in the city. Meanwhile the situation is getting worse in Mainland China. I chose the case study of Tsoi Yuen Resistance from 2008 to 2011 for revelation of the complicated colonial history and postcolonial era of Hong Kong. I focused on discussing the Tsoi Yuen Resistance and the Post-80s movement, and how they have exposed the tension between top-down urban planning and development and public movements fighting for a more democratic process in choosing their way of living. Through the study of a village movement which as well as the rationale behind the Post-80s’ support, I hoped to illustrate how this movement has awaken a different sense of living for the new generations in the midst of the high-sounding urban development. It is an opportunity to examine Hong Kong’s colonial epoch in a different perspective: through studying the Tsoi Yuen Village, let them (subalterns) speak for themselves. Furthermore, the significance of this resistance, taking place eleven years after the handover to the PRC, is an important fact that I shall not miss in later discussion. Last but not least, during the resistance, advanced technology and social networks such as Facebook, Twitter, iPhone were used by Post 80s generation to spread the latest information in order to attract public’s concern and participation. Therefore, apart from studying Tsoi Yuen Resistance as a local social movement, I also regard it as a part of the global movement in perusing ecological lifestyle and civil society. How Post 80s’ generation manipulates the global idea in a local context will also be examined.