162 resultados para Information processing
Resumo:
We consider the problem of structured classification, where the task is to predict a label y from an input x, and y has meaningful internal structure. Our framework includes supervised training of Markov random fields and weighted context-free grammars as special cases. We describe an algorithm that solves the large-margin optimization problem defined in [12], using an exponential-family (Gibbs distribution) representation of structured objects. The algorithm is efficient—even in cases where the number of labels y is exponential in size—provided that certain expectations under Gibbs distributions can be calculated efficiently. The method for structured labels relies on a more general result, specifically the application of exponentiated gradient updates [7, 8] to quadratic programs.
Resumo:
Spontaneous facial expressions differ from posed ones in appearance, timing and accompanying head movements. Still images cannot provide timing or head movement information directly. However, indirectly the distances between key points on a face extracted from a still image using active shape models can capture some movement and pose changes. This information is superposed on information about non-rigid facial movement that is also part of the expression. Does geometric information improve the discrimination between spontaneous and posed facial expressions arising from discrete emotions? We investigate the performance of a machine vision system for discrimination between posed and spontaneous versions of six basic emotions that uses SIFT appearance based features and FAP geometric features. Experimental results on the NVIE database demonstrate that fusion of geometric information leads only to marginal improvement over appearance features. Using fusion features, surprise is the easiest emotion (83.4% accuracy) to be distinguished, while disgust is the most difficult (76.1%). Our results find different important facial regions between discriminating posed versus spontaneous version of one emotion and classifying the same emotion versus other emotions. The distribution of the selected SIFT features shows that mouth is more important for sadness, while nose is more important for surprise, however, both the nose and mouth are important for disgust, fear, and happiness. Eyebrows, eyes, nose and mouth are important for anger.
Resumo:
We present an algorithm called Optimistic Linear Programming (OLP) for learning to optimize average reward in an irreducible but otherwise unknown Markov decision process (MDP). OLP uses its experience so far to estimate the MDP. It chooses actions by optimistically maximizing estimated future rewards over a set of next-state transition probabilities that are close to the estimates, a computation that corresponds to solving linear programs. We show that the total expected reward obtained by OLP up to time T is within C(P) log T of the reward obtained by the optimal policy, where C(P) is an explicit, MDP-dependent constant. OLP is closely related to an algorithm proposed by Burnetas and Katehakis with four key differences: OLP is simpler, it does not require knowledge of the supports of transition probabilities, the proof of the regret bound is simpler, but our regret bound is a constant factor larger than the regret of their algorithm. OLP is also similar in flavor to an algorithm recently proposed by Auer and Ortner. But OLP is simpler and its regret bound has a better dependence on the size of the MDP.
Resumo:
As the graphics race subsides and gamers grow weary of predictable and deterministic game characters, game developers must put aside their “old faithful” finite state machines and look to more advanced techniques that give the users the gaming experience they crave. The next industry breakthrough will be with characters that behave realistically and that can learn and adapt, rather than more polygons, higher resolution textures and more frames-per-second. This paper explores the various artificial intelligence techniques that are currently being used by game developers, as well as techniques that are new to the industry. The techniques covered in this paper are finite state machines, scripting, agents, flocking, fuzzy logic and fuzzy state machines decision trees, neural networks, genetic algorithms and extensible AI. This paper introduces each of these technique, explains how they can be applied to games and how commercial games are currently making use of them. Finally, the effectiveness of these techniques and their future role in the industry are evaluated.
Resumo:
Dealing with product yield and quality in manufacturing industries is getting more difficult due to the increasing volume and complexity of data and quicker time to market expectations. Data mining offers tools for quick discovery of relationships, patterns and knowledge in large databases. Growing self-organizing map (GSOM) is established as an efficient unsupervised datamining algorithm. In this study some modifications to the original GSOM are proposed for manufacturing yield improvement by clustering. These modifications include introduction of a clustering quality measure to evaluate the performance of the programme in separating good and faulty products and a filtering index to reduce noise from the dataset. Results show that the proposed method is able to effectively differentiate good and faulty products. It will help engineers construct the knowledge base to predict product quality automatically from collected data and provide insights for yield improvement.
Resumo:
Automatic species recognition plays an important role in assisting ecologists to monitor the environment. One critical issue in this research area is that software developers need prior knowledge of specific targets people are interested in to build templates for these targets. This paper proposes a novel approach for automatic species recognition based on generic knowledge about acoustic events to detect species. Acoustic component detection is the most critical and fundamental part of this proposed approach. This paper gives clear definitions of acoustic components and presents three clustering algorithms for detecting four acoustic components in sound recordings; whistles, clicks, slurs, and blocks. The experiment result demonstrates that these acoustic component recognisers have achieved high precision and recall rate.
Resumo:
The success of many knowledge-intensive industries depends on creative projects that lie at the heart of their logic of production. The temporality of such projects, however, is an issue that is insufficiently understood. To address this, we study the perceived time frame of teams that work on creative projects and its effects on project dynamics. An experiment with 267 managers assigned to creative project teams with varying time frames demonstrates that compared to creative project teams with a relatively longer time frame, project teams with a shorter time frame focus more on the immediate present, are less immersed in their task, and utilize a more heuristic mode of information processing. Furthermore, we find that time frame moderates the negative effect of team conflict on team cohesion. These results are consistent with our theory that the temporary nature of creative projects shapes different time frames among project participants, and that it is this time frame that is an important predictor of task and team processes.
Resumo:
This paper reports results from a study exploring the multimedia search functionality of Chinese language search engines. Web searching in Chinese (Mandarin) is a growing research area and a technical challenge for popular commercial Web search engines. Few studies have been conducted on Chinese language search engines. We investigate two research questions: which Chinese language search engines provide multimedia searching, and what multimedia search functionalities are available in Chinese language Web search engines. Specifically, we examine each Web search engine's (1) features permitting Chinese language multimedia searches, (2) extent of search personalization and user control of multimedia search variables, and (3) the relationships between Web search engines and their features in the Chinese context. Key findings show that Chinese language Web search engines offer limited multimedia search functionality, and general search engines provide a wider range of features than specialized multimedia search engines. Study results have implications for Chinese Web users, Website designers and Web search engine developers. © 2009 Elsevier Ltd. All rights reserved.
Resumo:
Several authors stress the importance of data’s crucial foundation for operational, tactical and strategic decisions (e.g., Redman 1998, Tee et al. 2007). Data provides the basis for decision making as data collection and processing is typically associated with reducing uncertainty in order to make more effective decisions (Daft and Lengel 1986). While the first series of investments of Information Systems/Information Technology (IS/IT) into organizations improved data collection, restricted computational capacity and limited processing power created challenges (Simon 1960). Fifty years on, capacity and processing problems are increasingly less relevant; in fact, the opposite exists. Determining data relevance and usefulness is complicated by increased data capture and storage capacity, as well as continual improvements in information processing capability. As the IT landscape changes, businesses are inundated with ever-increasing volumes of data from both internal and external sources available on both an ad-hoc and real-time basis. More data, however, does not necessarily translate into more effective and efficient organizations, nor does it increase the likelihood of better or timelier decisions. This raises questions about what data managers require to assist their decision making processes.
Resumo:
In the field of process mining, the use of event logs for the purpose of root cause analysis is increasingly studied. In such an analysis, the availability of attributes/features that may explain the root cause of some phenomena is crucial. Currently, the process of obtaining these attributes from raw event logs is performed more or less on a case-by-case basis: there is still a lack of generalized systematic approach that captures this process. This paper proposes a systematic approach to enrich and transform event logs in order to obtain the required attributes for root cause analysis using classical data mining techniques, the classification techniques. This approach is formalized and its applicability has been validated using both self-generated and publicly-available logs.
Resumo:
Complexity is a major concern which is aimed to be overcome by people through modeling. One way of reducing complexity is separation of concerns, e.g. separation of business process from applications. One sort of concerns are cross-cutting concerns i.e. concerns which are scattered and tangled through one of several models. In business process management, examples of such concerns are security and privacy policies. To deal with these cross-cutting concerns, the aspect orientated approach was introduced in the software development area and recently also in the business process management area. The work presented in this paper elaborates on aspect oriented process modelling. It extends earlier work by defining a mechanism for capturing multiple concerns and specifying a precedence order according to which they should be handled in a process. A formal syntax of the notation is presented precisely capturing the extended concepts and mechanisms. Finally, the relevant of the approach is demonstrated through a case study.
Resumo:
Privacy is an important component of freedom and plays a key role in protecting fundamental human rights. It is becoming increasingly difficult to ignore the fact that without appropriate levels of privacy, a person’s rights are diminished. Users want to protect their privacy - particularly in “privacy invasive” areas such as social networks. However, Social Network users seldom know how to protect their own privacy through online mechanisms. What is required is an emerging concept that provides users legitimate control over their own personal information, whilst preserving and maintaining the advantages of engaging with online services such as Social Networks. This paper reviews “Privacy by Design (PbD)” and shows how it applies to diverse privacy areas. Such an approach will move towards mitigating many of the privacy issues in online information systems and can be a potential pathway for protecting users’ personal information. The research has also posed many questions in need of further investigation for different open source distributed Social Networks. Findings from this research will lead to a novel distributed architecture that provides more transparent and accountable privacy for the users of online information systems.
Resumo:
Effective risk management is crucial for any organisation. One of its key steps is risk identification, but few tools exist to support this process. Here we present a method for the automatic discovery of a particular type of process-related risk, the danger of deadline transgressions or overruns, based on the analysis of event logs. We define a set of time-related process risk indicators, i.e., patterns observable in event logs that highlight the likelihood of an overrun, and then show how instances of these patterns can be identified automatically using statistical principles. To demonstrate its feasibility, the approach has been implemented as a plug-in module to the process mining framework ProM and tested using an event log from a Dutch financial institution.