Exploring Optimal Resource Utilization in the Evolving Data Resources

  • 張 健暐

Student thesis: Doctoral Thesis


In this thesis we discussed three significant aspects that need to be considered in the resource utilization of the evolving environment There are mainly three portions in this thesis including strengthening the duration of advertising impact optimizing medical resource placement to reduce hospital burden and improving resource allocation strategies in crowdsourcing First of all for the topic of strengthens the sustainability we study a novel paradigm of viral marketing with the goal to sustain the influential effectiveness in the network We study from real cases such as the Ice Bucket Challenges for the ALS awareness and figure out the ”easy come and easy go” phenomenon in the marketing promotion Such a natural property is fully unexplored in the literature but it will violate the need of many marketing applications which attempt to receive the perpetual attention and support We thus highlight the problem of Influential Sustainability to pursue the long-term and effective influence on the network Given the set of initial seeds S and a threshold ρ the goal of Influential Sustainability is to best decide the timing to activate each seed in S so as to maximize the number of iterations in which each iteration will activate the number of inactive nodes more than ρ The Influential Sustainability problem is challenging due to its #P-hard nature Secondly with the effects of global warming some epidemic diseases via mosquito (e g mosquito-borne diseases) become more serious such as dengue fever and zika virus It is reported that the epidemic disease may cause many challenges to the hospital management due to the unexpected burst with uncertain reasons Furthermore the imperfect cares during the propagation of epidemic diseases such as dengue fever (so far the appropriate treatment is not well established) may lead to the increasing mortality rate which should be avoided In this thesis a novel paradigm for optimizing the placement of medical resource is proposed in pursuit of reducing the overloading cases in hospitals during the epidemic outbreak in the urban area In this thesis we are the first thesis to explore two important issues including the strategy to evaluate the service quality and the solution to dynamically dispatch the medical resource along with the spatial variation of the epidemic outbreak As validated in our experimental results in real data of dengue outbreak happening in Tainan (2015) we present the feasibility of our framework to deploy a dynamic placement strategy for medical resource assignment Lastly researchers and scientists have been using crowdsourcing platforms to collect labeled training data in recent years The process is cost-effective and scalable but research has shown that the quality of truth inference is unstable due to worker bias work variance and task difficulty In this work we present a hybrid system that brings together a well-trained troop of domain experts and the multitudes of a crowdsourcing platform to collect high-quality training data for industry-level classification engines We show how to acquire high quality labeled data through quality control strategies that dynamically and cost-effectively leverage the strengths of both domain experts and crowdsourcing
Date of Award2018 Aug 8
Original languageEnglish
SupervisorKun-Ta Chuang (Supervisor)

Cite this