773 research outputs found
Enhancing a just transition finance system for carbon-intensive industries
No full textJunije Zhang, Fei Xie, Minjun Chang, Yufei Dai, Mengfu Ha
Climate change and its ecological and socioeconomic impact: evidence from China's historical document for Qing dynasty
No full textClimate change is currently of great concern by scientist and public. However, the ecological and socio-economic effects of long-term climate change remains largely unknown. To study on the long-term climate change, and its impacts on the environmental conditions, ecological and socio-economic consequences, the long time series of climate related events from historical records are needed to fill the gap of the instrumental data. In this study, a unique data source -- "The compendium of Chinese Meteorological records of the Last 3,000 years (Zhang 2004) which contains invaluable information about climate related events recorded in the China's historical documents were digitized for the last Dynasty. Pearson correlation test was conducted to test the relationship between crop harvest and climate events in case study one. The results revealed that climate conditions affected past agriculture harvest in China. Besides direct effect of cooling on the land carrying capacity, periodic ecological stresses such as drought events can significantly reduce the agricultural yield. The issue of stationarity of variables is the great concern in this study. Local variations both in temporal and spatial scale analysis were considered in the following case studies. Second case study applied continuous wavelet analysis for analyzing the local variation in temporal scale. The result revealed that the periodicity of fluctuations of locusts, temperature and drought series are consistently at around 100 year's band. The consistent associations between locust and temperature, temperature and drought, locust and drought at same frequency and time space indicted the possible casual interlinks of temperature-drought-locust plague. The finding suggests that drought events driven by long term variation of temperature change explains locust dynamics better than floods. The last case study used Geographically Weighted Regression methods for analyzing spatially varying relationships between determinate variable--famine and explanatory variables such as floods, droughts, poor harvest and locust outbreaks. The results implicated that all variables have significant effects on famine occurrence in last Dynasty of China. Among the explanatory variables, drought shows strongest effect on famine. The results also suggested that there are significant spatial variations across the study area. Therefore, it's important to consider the local regression methods for analyzing the relationships between famine and other climate conditions.Ph.D.Includes abstractVitaIncludes bibliographical referencesby Yufei Wan
The Experience of a Lifetime: Interactive Digital Experience Beyond the Screen
No full textScreen-based digital experience design is blooming among the local businesses in Metro Vancouver along with the increased pervasiveness of information technologies, new digital products in contemporary society. However, there are significantly fewer cases and related businesses around tangible interactive digital experience in which tangible objects and physical spaces replace the screen as the site of interaction. This thesis project aims to explore the specialties of the tangible interactive experience compared to the digital experience on the screen or in the virtual space. Additionally, the author investigates how to leverage user experience design methodologies in the process of designing an experimental interactive experience. In this practice-based exploration, the author prototyped four interactive digital experiences using different interactive technologies and tools tailored to different use case scenarios: 1. an interactive offline retail experience, 2. a “magical” and playful painting, 3. a room-scale interactive installation, and 4. an immersive meditation activity. These projects illustrate and explore the implementation of tangible interactions into digital experience design. During the development process, the author applied several user experience design methodologies in the projects – including field research, interviews, questionnaires, and design probes – to develop a workable framework designing tangible interactive experiences throughout the research project. The author aims to outline key implications of applying principles of user experience design to the field of tangible interactive environments. In the process, the author argues that tangible interactive design is indispensable in a successful and engaging digital experience, and thus worth investing in and exploring further in Vancouver’s marketplace
Trust under siege: Label spoofing attacks against machine learning for Android malware detection
No full textDesign of optimal investment policy for influence systems and online learning for job scheduling
No full textThis thesis contains two main research thrusts, one related to the design of optimal investment policies for influence systems and the other related to online learning for job scheduling on heterogeneous machines.
In Chapter 2, a continuous-time influence system on social networks is considered, where the goal is to decide on how to allocate a limited amount of budget over a finite time horizon in order to control the opinion of social entities towards a specific objective. Describing the model under continuous-time settings, it is shown that even under the simplistic case of one marketer and one social entity, the optimal strategy may not exist without any assumption on the number of investment switches. Several sufficient conditions are then developed, which guarantee the existence of an optimal policy. Subsequently, the structure of the optimal policy under some special cases are characterized.
In Chapter 3, a new model for online job scheduling on heterogeneous machines is developed. In that model, the goal is to schedule a sequence of arriving jobs on a set of heterogeneous machines in an online fashion with the overall quality of service as close as possible to an optimal offline benchmark. However, in practice, each machine may have an unknown different power/energy budget, and its welfare is proportional to the product of its power and its cumulative utilities. The goal is to minimize the regret, that is, the expected difference between the total quality of service (i.e., the sum of all the machines’ welfare) obtained by the algorithm and its maximum value had we known the power budgets a priori. First, it is shown that a simple Explore-then-Exploit scheduling algorithm achieves a sub-linear regret of O(T^{2/3}), where T is the total number of jobs. This result is then enhanced by providing an Upper Confidence Bound (UCB) algorithm achieving a logarithmic regret O(log T ).Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2022-08-01The student, Yufei Ruan, accepted the attached license on 2020-07-23 at 18:52.The student, Yufei Ruan, submitted this Thesis for approval on 2020-07-23 at 19:12.This Thesis was approved for publication on 2020-07-24 at 14:20.DSpace SAF Submission Ingestion Package generated from Vireo submission #15741 on 2020-10-02 at 15:34:10Made available in DSpace on 2020-10-07T22:44:48Z (GMT). No. of bitstreams: 3
RUAN-THESIS-2020.pdf: 522109 bytes, checksum: ce871cfc5c7476298d5c63b9ff5cef88 (MD5)
Yufei master thesis.zip: 3539520 bytes, checksum: e55ce38d00b60b6b6df355071f7029c6 (MD5)
LICENSE.txt: 4207 bytes, checksum: a30d881b5a08b0b8bb35bf129d458821 (MD5)
Previous issue date: 2020-07-24Embargo set by: Seth Robbins for item 116268
Lift date: 2022-10-07T22:44:53Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemAuthor requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Onl
- …
