Techniques for Artefact Conception and Development--论文代写范文精选

我们假设生物和文物都面对同样的基本困难，任何模型的现象是不完整的，总有一些隐藏变量影响这一现象。这些隐变量模型的影响,这种现象从来没有表现得完全一样。下面的essay代写范文讲述了这一问题。

Abstract
The purpose of this paper is to make a state of the art on probabilistic methodology and techniques for artefact conception and development. It is the 8th deliverable of the BIBA (Bayesian Inspired Brain and Artefacts) project. We first present the incompletness problem as the central difficulty that both living creatures and artefacts have to face: how can they perceive, infer, decide and act efficiently with incomplete and uncertain knowledge?. We then introduce a generic probabilistic formalism called Bayesian Programming. This formalism is then used to review the main probabilistic methodology and techniques. This review is organized in 3 parts: first the probabilistic models from Bayesian networks to Kalman filters and from sensor fusion to CAD systems, second the inference techniques and finally the learning and model acquisition and comparison methodologies. We conclude with the perspectives of the BIBA project as they rise from this state of the art. 
Keywords: Bayesian programming, Bayesian modelling, Bayesian reasoning, Bayesian learning, Bayesian networks

INCOMPLETENESS AND UNCERTAINTY 
We think that over the next decade, probabilistic reasoning will provide a new paradigm for understanding neural mechanisms and the strategies of animal behaviour at a theoretical level, and will raise the performance of engineering artefacts to a point where they are no longer easily outperformed by the biological examples they are imitating. The BIBA project has been motivated by this conviction and aims to advance in this direction. We assume that both living creatures and artefacts have to face the same fundamental difficulty: incompleteness (and its direct consequence uncertainty). Any model of a real phenomenon is incomplete: there are always some hidden variables, not taken into account in the model, that influence the phenomenon. The effect of these hidden variables is that the model and the phenomenon never behave exactly alike. Both living organisms and robotic systems must face this central difficulty: how to use an incomplete model of their environment to perceive, infer, decide and act efficiently? Rational reasoning with incomplete information is quite a challenge for artificial systems. 

The purpose of probabilistic inference and learning is precisely to tackle this problem with a well-established formal theory. During the past years a lot of progress has been made in this field both from the theoretical and applied point of view. The purpose of this paper is to give an overview of these works and especially to try a synthetic presentation using a generic formalism named Bayesian Programming (BP) to present all of them. It is not an impartial presentation of this field. Indeed, we present our own subjective "subjectivist" point of view and we think that is why it may be interesting. The paper, after discussing further the incompleteness and uncertainty problems and how probabilistic reasoning helps deal with them, is organised in 4 main parts. The first part is a short and formal presentation of BP. The second part is a presentation of the main probabilistic models found in the literature. The third part describes the principal techniques and algorithms for probabilistic inference. Finally, the fourth deals with the learning aspects.

The programmer uses this abstract representation to program the robot. The programs use these geometric, analytic and symbolic notions. In a way, the programmer imposes on the robot his or her own abstract conception of the environment. The difficulties of this approach appear when the robot needs to link these abstract concepts with the raw signals it obtains from its sensors and sends to its actuators. The central origin of these difficulties is the irreducible incompleteness of the models. Indeed, there are always some hidden variables, not taken into account in the model, that influence the phenomenon. The effect of these hidden variables is that the model and the phenomenon never behave exactly the same. The hidden variables prevent the robot from relating the abstract concepts and the raw sensory-motor data reliably. The sensory-motor data are then said to be «noisy» or even «aberrant». 

An odd reversal of causality occurs that seem to consider that the mathematical model is exact and that the physical world has some unknown flaws. Controlling the environment is the usual answer to these difficulties. The programmer of the robot looks for the causes of «noises» and modifies either the robot or the environment to suppress these «flaws». The environment is modified until it corresponds to its mathematical model. This approach is both legitimate and efficient from an engineering point of view. A precise control of both the environment and the tasks ensures that industrial robots work properly. However, compelling the environment may not be possible when the robot must act in an environment not specifically designed for it. In that case, completely different solutions must be devised.

A tantalising answer is to say that natural evolution provided living beings with both the pertinent variables and the adequate decomposition and parametric forms. The pertinent variables may have been obtained by selecting the sensors and actuators in order to supply vital information. The decomposition would correspond to the structure of the nervous system, which basically expresses dependencies and conditional independencies between variables. The parametric forms can be seen as the information processing units implemented by neurons and assembly of neurons. Given this apparatus, correponding to the preliminary knowledge, each individual in its lifetime can answer the first question by experimenting and learning the values of the free parameters of its nervous system.（essay代写)

51Due网站原创范文除特殊说明外一切图文著作权归51Due所有；未经51Due官方授权谢绝任何用途转载或刊发于媒体。如发生侵犯著作权现象，51Due保留一切法律追诉权。（essay代写)
更多essay代写范文欢迎访问我们主页 www.51due.com 当然有essay代写需求可以和我们24小时在线客服 QQ:800020041 联系交流。-X（essay代写)