Recommendation systems are emerging as an important business application as the demand for personalized services in E-commerce increases. Collaborative filtering techniques are widely used for predicting a user's preference or generating a list of items to be recommended. In this thesis, we develop several new approaches for collaborative filtering based on model combination and kernel smoothing. Specifically, we start with an experimental study that compares a wide variety of CF methods under different conditions. Based on this study, we formulate a combination model similar to boosting but where the combination coefficients are functions rather than constant. In another contribution we formulate and analyze a local variation of matrix factorization. This formulation constructs multiple local matrix factorization models and then combines them into a global model. This formulation is based on the local low-rank assumption, a slightly different but more plausible assumption about the rating matrix. We apply this assumption to both rating prediction and ranking problems, with both empirical validations and theoretical analysis. We contribute with this thesis in four aspects. First, the local approaches we present significantly improve the accuracy of recommendations both in rating prediction and ranking problems. Second, with the more realistic local low-rank assumption, we fundamentally change the underlying assumption for matrix factorization-based recommendation systems. Third, we present highly efficient and scalable algorithms which take advantage of parallelism, suited for recent large scale datasets. Lastly, we provide an open source software implementing the local approaches in this thesis as well as many other recent recommendation algorithms, which can be used both in research and production.

Mining has a long history of production and operation management. Economies of scales have changed drastically and technology has transformed the mining industry significantly. One of the most important technological improvements is increased equipment, human, and plant tracking capabilities. This provided a continuous data stream to the decision makers, considering dynamic operational conditions. However, managerial approaches did not change in parallel. Even though many process improvement tools using equipment/human/plant tracking capabilities were developed (Fleet Management Systems, Plant Monitoring Systems, Workforce Management Systems etc.), to date there is no holistic approach or system to manage the entire value chain in mining. Mining operations are designed and managed around the already known system designated bottlenecks. However, contrary to common belief in mining, bottlenecks are not static. They can shift from one process or location to another. It is important for management to be aware of the new bottlenecks, since their decisions will be effected. Therefore, identification of true bottlenecks in real-time will help tactical level decisions (use of buffers, resource transfer), and identification of historical bottlenecks will help strategic-level decisions (investments, increasing capacity etc.). This thesis aims to address the managerial focus on the true bottlenecks. This is done by first identifying and ranking true bottlenecks in the system. The study proposes a methodology for creating Bottleneck Identification Model (BIM) that can identify true bottlenecks in a value chain in real-time or historically, depending on the available data. This approach consists of three phases to detect and rank the bottlenecks. In the first phase, the system is defined and variables are identified. In the second phase, the capacity, rates, and buffers are computed. In the third phase, considering particularities of the mine exceptions are added by taking mine characteristics into account, and bottlenecks are identified and ranked.

No description available.

Hazard communication is an essential part of a comprehensive safety plan, especially for those facilities that contain reactive chemicals. There are a variety of means of communicating a chemical hazard, but one of the most prevalent in the United States is the Instability Rating found in the NFPA 704 standard. While the NFPA 704 identifies hazards associated with exothermically decomposing compounds, it neglects compounds that decompose endothermicly to form large quantities of gas. Such compounds have been known to cause accidents due to pressure buildup, such as in the BP Amoco Polymers explosion in 2001. In this work, twenty-five compounds were examined via an APTAC to determine their pressure and temperature profiles. These profiles were then used to determine the amount of gas generated, the gas generation rate, the gas generation product, the onset temperature, and the instantaneous power density. These properties were analyzed to determine those that best represented the instability hazard of the chemical. Ultimately, the molar gas generation rate and onset temperature were chosen to rate the selected chemicals, and new cut-offs were established to divide the chemicals into revised instability groupings. Compounds that did not decompose in the temperature range examined were given the rating of zero. Compounds with low onset temperatures and high gas generation rates were assigned the rating of 4, while chemicals with high T_onset and low dn/dt_maxn were assigned a value of 1. Chemicals with high onset temperatures and high gas generation rates were grouped into rating 3. Group 2 included low onset temperature compounds with low gas generation rates. The cut-offs used to define these regions were 130 degrees C for the onset temperature and 0.01 (1/min) for the gas generation rate. The ratings were found to be comparable to the current NFPA system, but improved upon it by providing a valid rating (group 1) for the chemicals that endothermically generated gas. Detailed plots of the data are provided as well as suggestions for future work.

This dissertation presents three different closely related topics on the value of eco-friendliness in the financial market. The first essay attempts to estimate hedonic stock price model to find a contemporaneous relationship between stock return and firms’ environmental performance and recover the value of investor’s willingness to pay of eco-friendliness. This study follows stock and environmental performances of the 500 largest US firms from 2009 to 2012. The firms’ environmental data come from the Newsweek Green Ranking, both aggregate measures: green ranking (GR) and green score (GS), and disaggregate measures: environmental impact score (EIS), green policy and performance score (GPS), reputation survey score (RSS), and environmental disclosure score (EDS). The results show a non-linear relationship between environmental variables and stock return, i.e. upside down bowl shape or increasing in decreasing rate. That means for low green ranking firms the marginal effect is positive while for high green ranking firms the marginal effect is negative. The investor’s willingness to pay (WTP) for a greener stock for firms in the lowest 25 green ranking, on average, is 0.0096% higher stock price. The second essays attempt to determine if a firm’s environmental performance affects future systematic risk. Systematic risk measures an individual stock’s volatility relative to the market price. This study also uses the Newsweek Green Ranking’s environmental variables. The results show significant evidence of a non-linear relationship between green variables and systematic (market) risk, but the shape is not unanimous for all environmental variables. The shape of the relationship for green ranking (GR), for example, is U-shape. This means that for the firms in the bottom rank, improving rank will lower systematic (market) risk, and for the firms in the top rank improving rank will increase systematic (market) risk. On average the marginal effect for the firms in the bottom and top 25 firms are -0.2% and 0.09% respectively. The third essay is the effect of a firm’s environmental performances on a firm’s idiosyncratic risk. Idiosyncratic risk measures an individual stock’s volatility independent from the market price. This study also uses the Newsweek Green Ranking’s environmental variables. The results show significant non-linear relationships between environmental variables and idiosyncratic risk, even though there is no unanimous shape among the environmental variables. In the case of green ranking, for example, it has U-shape; for the firms in the bottom rank, improving green ranking will lower idiosyncratic risk and for firm in the top green ranking, improving green ranking will increase idiosyncratic risk. On average the marginal effect for firm in bottom and top 25 firms are -0.4% and 0.2% respectively.

The traditional way to analyze stocks and portfolios within the area of finance have been restricted to Sharpe and Markovitz. The Omega function and its properties enlighten the field of finance and differs from the traditional ways when it comes to the volatility of the stocks. The Omega function, the Sharpe performance criteria and mean-variance model by Markovitz will be used. All calculations are done in Matlab and the data sheets are excel tables. The aim of this thesis is to investigate the nordic small cap market by using the Omega function, Sharpe performance criteria and the mean variance model by Markovitz. In order to to see how the purposed methods differs.

In this study, Turkey&rsquo / s Internet visibility will be analyzed based on data to be collected from multiple different resources (such as / Google, Yahoo, Altavista, Bing and AOL). Analysis work will involve inspection of DNS queries, Web crawling and some other similar techniques. Our goal is to investigate global Internet and find webs that has common pattern of representing Internet visibility of Turkey and compare their characteristics with other webs&#039 / on the world and discover their similarities and differences.

Face recognition can be used in various biometric applications ranging from identifying criminals entering an airport to identifying an unconscious patient in the hospital With the introduction of 3-dimensional scanners in the last decade, researchers have begun to develop new methods for 3-D face recognition. This thesis focuses on 3-D face recognition using the one- and two-dimensional Discrete Cosine Transform (DCT) . A feature ranking based dimensionality reduction strategy is introduced to select the DCT coefficients that yield the best classification accuracies. Two forms of 3-D representation are used: point cloud and depth map images. These representations are extracted from the original VRML files in a face database and are normalized during the extraction process. Classification accuracies exceeding 97% are obtained using the point cloud images in conjunction with the 2-D DCT.

In a dynamically monitored environment the analyst team need timely and accurate information to conduct proactive action over complex situations. Typically, there are thousands of reported activities in a real time operation, therefore steps are taken to direct the analyst’s attention to the most important activity. The data fusion community have introduced the information fusion model, with multiple situational assessments. Each process lends itself to ranking the most important activities into a predetermined order. Unfortunately, the capability of a real time system can be hindered by the knowledge limitation problem, particularly when the underlying system is processing multiple sensor information. Consequently, the situational awareness domains may not rank the identified situation as perfect, as desired by the decision-making resources. This thesis presents advanced research carried out to evaluate the ranking capability of information from the situational awareness domains: perception, comprehension and projection. The Ranking Capability Score (RCS) has been designed for evaluating the prioritisation process. The enhanced (RCS) has been designed for addressing the knowledge representation problem in the user system relation under a situational assessment where the proposed number of tracking activities are dynamically shifted. Finally, the Scheduling Capability Score was designed for evaluating the scheduling capability of the situational awareness system. The proposed performance metrics have been successful in fulfilling their objectives. Furthermore, they have been validated and evaluated using an analytical approach, through conducting a rigorous analysis of the prioritisation and scheduling processes, despite any constraints related to a domain-specific configuration.

Background and context: Given the constant evolving developments in information technology (IT) in healthcare in Canada and Ontario, and the relatively nonexistent body of literature on IT management issues from the perspectives of top IT managers (i.e. Chief Information Officers, IT directors, IT top managers) in hospitals, a follow up study of IT management issues to the study done by Jaana et al. is conducted. Purpose: To develop an authoritative list of IT management issues in Ontario hospitals and compare the results to the earlier study and the literature. Methods: Using the Ranking Type Delphi technique, the responses from IT top managers in three main panels of Ontario hospitals were solicited through a controlled iterative feedback process. The hospitals were divided into the academic panel (n = 6), community panel (n = 12), and the rural panel (n = 8) for a total of 26 out of 33 participants who completed the study. Results: 26 issues were raised and a total of 24 issues were ranked in the study. Among the 14 common issues between the three panels, the top five issues were limited funding, keeping infrastructure current, external security threats, increasing cost, and managing demands for IT projects. Comparing with the study by Jaana et al. (2011), a total of 7 new issues emerged which are concerned with technology, regulatory challenges, and human issues. A total of 10 issues were dropped from the earlier study spanning areas of strategic, technological, organizational, and human issues. The participants in the study did not significantly differ individually based on their background characteristics, where the only significant difference observed between the hospital panels was due to hospital characteristics. During the brainstorming phase a total of 195 issues were provided which were consolidated by two researchers to form a list of 26 IT management issues, with an inter coder reliability of 88%. The issues with a 4.5 out of 7 rating and higher on a Likert scale were retained to narrow down the list. This resulted in 19 issues for the rural and community panels, and 21 issues for the academic panel, with 14 of the 26 issues being common to all three panels. The ranking phase was conducted with two rounds of ranking due to the low consensus levels during the first round. The consensus level after two rounds was; W academic = 0.235, W community = 0.254, and W rural = 0.381. Contributions: This study presents a significant contribution to the management of medical informatics field by providing an approach to categorize IT management issues to observe trends overtime as well as present the application of a seminal framework to explain the changes in these issues as organizations change and grow overtime. At the management and practical levels, the list of prioritized issues provides an evidence base for top IT managers to make IT related decisions at the organizational level. The list also acts as a second benchmarking tool to evaluate hospital performance overtime with the various issues. At the policy development level, provincial governments can use the list to devise comprehensive IT management strategies to address the various regulatory issues reported. Future research can focus on exploring the resonating behind the rankings provided and replicating this study over time and across various geographies so that a large survey can be developed to follow the evolutions of IT management issues in healthcare over time.