This master thesis in geography investigates how remote sens- ing can be used in Transnational Corporations (TNC) global Corporate Social Responsibility (CSR) initiatives. The study aims to delineate an accurate method in remote sensing to be used to monitor deforestation in global value chains. Research questions asked are 1) What are the current monitoring practises used by TNCs to monitor global value chains? 2) Which is the most user-friendly and accurate remote sensing technique to map deforestation? 3) How can remote sensing successfully be implemented in TNCs CSR-initiatives? The study is approached from two perspectives, building on theories of value chains, and qualitative methods to answer the first research question. While the second question is a method study, investigating how well a spectral approach versus a contextual approach can map deforest- ation in Landsat scenes. The results are compared with Global Forest Watch (GFW), and the highest accuracy were acquired from the WICS (Window Indipendent Context Segmentation) technique. Conclusions includes that remote sensing can be used in CSR initiatives, to establish a baseline level or as a fifth dimen- sion in a score sheet approach. However, inconclusive mapping of value chains are a big hinder today.

The aim of this study is to compare remote sensing methods in the context of a vegetation mapping of an urban environment. The methods used was (1) a traditional per-pixel based method; maximum likelihood supervised classification (ENVI), (2) a standard object based method; example based feature extraction (ENVI) and (3) a newly developed method; Window Independent Contextual Segmentation (WICS) (Choros Cognition). A four-band SPOT5 image with a pixel size of 10x10m was used for the classifications. A validation data-set was created using a ortho corrected aerial image with a pixel size of 1x1m. Error matrices was created by cross-tabulating the classified images with the validation data-set. From the error matrices, overall accuracy and kappa coefficient was calculated. The object-based method performed best with a overall accuracy of 80% and a kappa value of 0.6, followed by the WICS method with an overall accuracy of 77% and a kappa value of 0.53, placing the supervised classification last with an overall accuracy of 71% and a kappa value of 0.38. The results of this study suggests object-based method and WICS to perform better than the supervised classification in an urban environment.

領導者之決策對於組織的影響甚大，不僅決定企業的獲利，更可能影響組織的存亡。在現今快速變遷的環境中，領導者經常面臨在時間壓力下需快速做出正確的決策，因此本研究著重於探討領導者才能對於領導者決策品質，以及團隊競爭力與績效之影響。具體而言，本研究從Sternberg所提出的WICS模式為出發點，探討「領導者智慧」、「領導者實用智力」、「領導者分析智力」及「領導者創新能力」對於「領導者決策品質」、「團隊競爭力與績效」之影響，並建構「領導者才能之競爭力模式」。至於領導者才能之競爭力模式，不僅可以做為後續相關研究之重要參考，亦可協助企業與領導者檢視其領導效能，進而做為改善競爭力與績效之重要依據。 本研究之問卷經嚴謹之程序編製，首先進行專家會議、信效度及因素分析之檢驗，之後再進行研究變項間之探討。研究樣本採立意取樣方式，以團隊為單位，每個團隊收集團隊領導者及其五位直屬部屬之調查問卷。本研究總計回收40位領導者與200位部屬之有效樣本，統計分析結果顯示，本研究所有量表皆具有良好之內部一致性信度（.754~.972）與效標關聯效度（.280~.820）。 此外，本研究根據結構方程模式驗證本研究之假設，綜合結果顯示：1. 領導者智慧、實用智力、分析智力及創新能力對於領導者決策品質具有顯著且正向之影響；2. 領導者決策品質對於團隊競爭力具有顯著且正向之影響；3. 團隊競爭力對於團隊績效具有顯著且正向之影響；4. 「領導者才能之競爭力模式」獲得實證支持。最後，研究者根據本研究結果，提出對未來學術研究與實務應用具體之建議。 / Leaders’ decisions have a great impact on organizations; they influence not only the revenue but also the survival of organizations. In this fast-changing environment, leaders always need to make fast and correct decisions under time pressure. Therefore, the focus of this study is the effect of leaders’ skills on the quality of leaders’ decisions so as to improve the competitiveness and performance of teams or organizations. The present study starts from WICS model which developed by Sternberg to examine the effect of leaders’ wisdom, practical intelligence, analytic intelligence, and creativity on the quality of leaders’ decisions, team competitiveness and performance, and to construct the competitiveness model of leaders’ skills. The development of competitiveness model of leaders’ skills could not only be used by following related studies, but also assist organizations and leaders to review leadership efficacy in order to improve the competitiveness and performance. The present study constructed the questionnaire through an expert committee, tests of reliability, validity, and factor analysis. The samples of team unit were collected using purposive sampling, and each unit included one leader and five immediate subordinates. A total of forty valid sample units which had forty leaders and two hundred immediate subordinates were collected. Through the statistical analysis, it shows the internal consistency reliability (.752-.972) and criterion-related validity (.280-.820) of all scales (leaders’ self-rating and subordinates’ other-rating) were excellent. Furthermore, according to the result of hypothesis testing and Structural Equation Modeling analysis, it shows: 1. Leaders’ wisdom, empirical intelligence, analytic intelligence, and creativity have a significant and positive influence on the quality of leaders’ decisions; 2. The quality of leaders’ decisions has a significant and positive influence on team competitiveness; 3. Team competitiveness has a significant and positive influence on team performance; 4. The constructed competitiveness model of leaders’ skills was supported. Finally, suggestions were proposed for further academic studies and practical application based on the result of the present study.

Wireless Sensor Networks (WSNs) consist of a large number of resource constrained sensor nodes equipped with various sensing devices which can monitor events in the real world. There are various applications such as environmental monitoring, target tracking forest fire detection, etc., which require clock synchronization among the sensor nodes with certain accuracy. However, a major constraint in the design of clock synchronization protocols in WSNs is that sensor nodes of WSNs have limited energy and computing resources. Clock synchronization process in the WSNs is carried out at each sensor node either synchronously, i.e., periodically during the same real-time interval, which we call synchronization phase, or asynchronously, i.e., independently without worrying about what other nodes are doing for clock synchronization. A disadvantage of asynchronous clock synchronization protocols is that they require the sensor nodes to remain awake all the time. Therefore, they cannot be integrated with any sleep-wakeup scheduling scheme of sensor nodes, which is a major technique to reduce energy consumption in WSNs. On the other hand, synchronous clock synchronization protocols can be easily integrated with the synchronous sleep-wakeup scheduling scheme of sensor nodes, and at the same time, they can provide support to achieve sleep-wakeup scheduling of sensor nodes. Essentially, there are two ways to synchronize the clocks of a WSN, viz. internal clock synchronization and external clock synchronization. The existing approaches to internal clock synchronization in WSNs are mostly hop-by-hop in nature, which is difficult to maintain. There are also many application scenarios where external clock synchronization is the only option to synchronize the clocks of a WSN. Besides, it is also desired that the internal clock synchronization protocol used is fault-tolerant to message loss and node failures. Moreover, when the external source fails or reference node fails, the external clock synchronization protocol should revert back to internal clock synchronization protocol with/without using any reference node. Towards this goal, first we propose three fully distributed synchronous clock synchronization protocols, called Energy Efficient and Fault-tolerant Clock Synchronization (EFCS) protocol, Weighted Average Based Internal Clock Synchronization (WICS) protocol, and Weighted Average Based External Clock Synchronization (WECS) protocol, for WSNs making use of peer-to-peer approach. These three protocols are dynamically interchangeable depending upon the availability of external source or reference nodes. In order to ensure consistency of the synchronization error in the long run, the neighboring nodes need to be synchronized with each other at about the same real time, which requires that the synchronization phases of the neighboring nodes always overlap with each other. To realize this objective, we propose a novel technique of pullback, which ensures that the synchronization phases of the neighboring nodes always overlap. In order to further improve the synchronization accuracy of the EFCS, WICS, and WECS protocol, we have proposed a generic technique which can be applied to any of these protocols, and the improved protocols are referred as IEFCS, IWICS, and IWECS respectively. We then give an argument to show that the synchronization error in the improved protocols is much less than that in the original protocols. We have analyzed these protocols for bounds on synchronization error, and shown that the synchronization error is always upper bounded. We have evaluated the performance of these protocols through simulation and experimental studies, and shown that the synchronization accuracy achieved by these protocols is of the order of a few clock ticks even in very large networks. The proposed protocols make use of estimated drift rate to provide logical time from the physical clock value at any instant and at the same time ensure the monotonicity of logical time even though physical clock is updated at the end of each synchronization phase. We have also proposed an energy aware routing protocol with sleep scheduling, which can be integrated with the proposed clock synchronization protocols to reduce energy consumption in WSNs further.