模式識別中英文

1、The Science of Pattern RecognitionAchievements and PerspectivesRobert P.W. Duin1 and Elzbieta P_ekalska21 ICT group, Faculty of Electr.Eng., Mathematics and Computer ScienceDelftUniversityofTechnology, TheN2SchoolofComputerScience,University of Manchester,United Kingdompekals

2、kacs.man.ac.ukSummary.Automatic pattern recognition is usually considered as an engineering area which focusses on the development and evaluation of systems that imitate or assist humans in their ability of recognizing patterns. It may, however, also be considered as a science that studies the facul

3、ty of human beings (and possibly other biological systems) to discover, distinguish, characterize patterns in their environment and accordingly identify new observations. The engineering approach to pattern recognition is in this view an attempt to build systems that simulate this phenomenon. By doi

4、ng that, scientific understanding is gained of what is needed in order to recognize patterns, in general.Like in any science understanding can be built from different, sometimes even opposite viewpoints. We will therefore introduce the main approaches to the science of pattern recognition as two dic

5、hotomies of complementary scenarios. They give rise to four different schools, roughly defined under the terms of expert systems, neural networks, structural pattern recognition and statistical pattern recognition.We will briefly describe what has been achieved by these schools, what is common and w

6、hat is specific, which limitations are encountered and which perspectives arise for the future. Finally, we will focus on the challenges facing pattern recognition in the decennia to come. They mainly deal with weaker assumptions of the models to make the corresponding procedures for learning and re

7、cognition wider applicable. In addition, new formalisms need to be developed.IntroductionWe are very familiar with the human ability of pattern recognition. Since our early years we have been able to recognize voices, faces, animals, fruits or inanimate objects. Before the speaking faculty is develo

8、ped, an object like a ball is recognized, even if it barely resembles the balls seen before. So, except for the memory, the skills of abstraction and generalization are essential to find our way in the world. In later years we are able to deal with much more complex patterns that may not directly be

9、 based on sensorial observations.For example, we can observe the underlying theme in a discussion or subtle patterns in human relations. The latter may become apparent, e.g. only by listening to somebodys complaints about his personal problems at work that again occur in a completely new job. Withou

10、t a direct participation in theevents, we are able to see both analogy and similarity in examples as complex as social interaction between people. Here, we learn to distinguish the pattern from just two examples.The pattern recognition ability may also be found in other biological systems:the cat kn

11、ows the way home, the dog recognizes his boss from the footsteps or the bee finds the delicious flower. In these examples a direct connection can be made to sensory experiences. Memory alone is insufficient; an important role is that of generalization from observations which are similar,although not

12、 identical to the previous ones. A scientific challenge is to find out how this may work.Scientific questions may be approached by building models and, more explicitly, by creating simulators, i.e. artificial systems that roughly exhibit the same phenomenon as the object under study. Understanding w

13、ill be gained while constructing such a system and evaluating it with respect to the real object. Such systems may be used to replace the original ones and may even improve some of their properties. On the other hand, they may also perform worse in other aspects. For instance, planes fly faster than

14、 birds but are far from being autonomous. We should realize, however, that what is studied in this case may not be the bird itself, but more importantly, the ability to fly.Much can be learned about flying in an attempt to imitate the bird, but also when differentiating from its exact behavior or ap

15、pearance. By constructing fixed wings instead of freely movable ones, the insight in how to fly grows.Finally, there are engineering aspects that may gradually deviate from the original scientific question. These are concerned with how to fly for a long time, with heavy loads, or by making less nois

16、e, and slowly shift the point of attention to other domains of knowledge.The above shows that a distinction can be made between the scientific study of pattern recognition as the ability to abstract and generalize from observations and the applied technical area of the design of artificial pattern r

17、ecognition devices without neglecting the fact that they may highly profit from each other. Note that patterns can be distinguished on many levels,starting from simple characteristics of structural elements like strokes, through features of an individual towards a set of qualities in a group of indi

18、viduals,to a composite of traits of concepts and their possible generalizations. A pattern may also denote a single individual as a representative for its population, model or concept. Pattern recognition deals, therefore, with patterns, regularities,characteristics or qualities that can be discusse

19、d on a low level of sensory measurements (such as pixels in an image) as well as on a high level of the derived and meaningful concepts (such as faces in images). In this work, we will focus on the scientific aspects, i.e. what we know about the way pattern recognition works and, especially, what ca

20、n be learned from our attempts to build artificial recognition devices.A number of authors have already discussed the science of pattern recognition based on their simulation and modeling attempts. One of the first, in the beginning of the sixties, was Sayre 64, who presented a philosophical study o

21、n perception, pattern recognition and classification. He made clear that classification is a task that can be fulfilled with some success, but recognition either happens or not. We can stimulate the recognition by focussing on some aspects of the question. Although we cannot set out to fully recogni

22、ze an individual, we can at least start to classify objects on demand. The way Sayre distinguishes between recognition and classification is related to the two subfields discussed in traditional texts on pattern recognition, namely unsupervised and supervised learning. They fulfill two complementary

23、 tasks. They act as automatic tools in the hand of a scientist who sets out to find the regularities in nature.Unsupervised learning(also related to exploratory analysis or cluster analysis) gives the scientist an automatic system to indicate the presence of yet unspecified patterns (regularities) i

24、n the observations. They have to be confirmed (verified) by him. Here, in the terms of Sayre, a pattern is recognized.Supervised learningis an automatic system that verifies (confirms)the patterns described by the scientist based on a representation defined by him. This is done by an automatic class

25、ification followed by an evaluation.In spite of Sayres discussion, the concepts of pattern recognition and classification are still frequently mixed up. In our discussion, classification is a significant component of the pattern recognition system, but unsupervised learning may also play a role ther

26、e. Typically, such a system is first presented with a set of known objects, the training set, in some convenient representation. Learning relies on finding the data descriptions such that the system can correctly characterize, identify or classify novel examples. After appropriate preprocessing and

27、adaptations, various mechanisms are employed to train the entire system well. Numerous models and techniques are used and their performances are evaluated and compared by suitable criteria. If the final goal is prediction, the findings are validated by applying the best model to unseen data. If the

28、final goal is characterization, the findings may be validated by complexity of organization (relations between objects) as well as by interpretability of the results.Fig. 1 shows the three main stages of pattern recognition systems: Representation, Generalization and Evaluation, and an intermediate

29、stage of Adaptation20. The system is trained and evaluated by a set of examples, the Design Set. The components are:Design Set.It is used both for training and validating the system. Given the background knowledge, this set has to be chosen such that it is representative for the set of objects to be

30、 recognized by the trained system.There are various approaches how to split it into suitable subsets for training,validation and testing. See e.g. 22, 32, 62, 77 for details.Representation.Real world objects have to be represented in a formal way in order to be analyzed and compared by mechanical me

31、ans such as a computer. Moreover, the observations derived from the sensors or other formal representations have to be integrated with the existing, explicitly formulated knowledge either on the objects themselves or on the class they may belong to. The issue of representation is an essential aspect

32、 of pattern recognition and is different from classification. It largely influences the success of the stages to come.Adaptation.It is an intermediate stage between Representation and Generalization,in which representations, learning methodology or problem statement are adapted or extended in order

33、to enhance the final recognition.This step may be neglected as being transparent, but its role is essential.It may reduce or simplify the representation, or it may enrich it by emphasizing particular aspects, e.g. by a nonlinear transformation of features that simplifies the next stage. Background k

34、nowledge may appropriately be (re)formulated and incorporated into a representation. If needed, additional representations may be considered to reflect other aspects of the problem. Exploratory data analysis (unsupervised learning) may be used to guide the choice of suitable learning strategies.Gene

35、ralization or Inference.In this stage we learn a concept from a training set, the set of known and appropriately represented examples, in such a way that predictions can be made on some unknown properties of new examples. We either generalize towards a concept or infer a set of general rules that de

36、scribe the qualities of the training data. The most common property is the class or pattern it belongs to, which is the above mentioned classification task.Evaluation.In this stage we estimate how our system performs on known training and validation data while training the entire system. If the resu

37、lts are unsatisfactory, then the previous steps have to be reconsidered.Different disciplines emphasize or just exclusively study different parts of this system. For instance, perception and computer vision deal mainly with the representation aspects 21, while books on artificial neural networks 62,

38、machine learning 4, 53 and pattern classification 15 are usually restricted to generalization. It should be noted that these and other studies with the words “pattern” and “recognition” in the title often almost entirely neglect the issue of representation. We think, however, thatthe main goal of th

39、e field of pattern recognition is to study generalization in relation to representation20.In the context of representations, and especially images, generalization has been thoroughly studied by Grenander 36. What is very specific and worthwhile is that he deals with infinite representations (say, un

40、sampled images),thereby avoiding the frequently returning discussions on dimensionality and directly focussing on a high, abstract level of pattern learning. We like to mention two other scientists that present very general discussions on the pattern recognition system: Watanabe 75 and Goldfarb 31,

41、32. They both emphasize the structural approach to pattern recognition that we will discuss later on. Here objects are represented in a form that focusses on their structure.A generalization over such structural representations is very difficult if one aims to learn theconcept, i.e. the underlying,

42、often implicit definition of a pattern class that is able to generate possible realizations. Goldfarb argues that traditionally used numeric representations are inadequate and that an entirely new, structural representation is necessary. We judge his research program as very ambitious, as he wants t

43、o learn the (generalized) structure of the concept from the structures of the examples. He thereby aims to make explicit what usually stays implicit. We admit that a way like his has to be followed if one ever wishes to reach more in concept learning than the ability to name the right class with a h

44、igh probability, without having built a proper understanding.模式識別研究的成果與展望 自動模式識別通常被認為是這樣的一個工程領域：專注于開發(fā)和評價模仿或輔助人類識別模式能力的系統(tǒng)，但是也可能被認為是這樣的一門科學：學習人類(或其它生物系統(tǒng))在所處環(huán)境中發(fā)現(xiàn)、區(qū)別和找出特征從而標識出觀察結果的本領。模式識別中工程的觀點是試圖建立模擬生物識別能力的系統(tǒng)，通過工程中的實踐，總的來說，科學上的理解在模式識別中的技術需求方面得到了發(fā)展。 象任何科學一樣，對模式識別的理解能夠從不同方向來建立，有時甚至是相反的觀點。我們將介紹模式識別科學中的主要

45、方法，即兩種不同方向且各有兩個不同種類的技術，這些技術產(chǎn)生了四個不同學派，粗略地可以定義為：專家系統(tǒng)，神經(jīng)網(wǎng)絡，結構模式識別和統(tǒng)計模式識別。 我們將簡要地描述這四個學派的發(fā)展成果，它們之間的相同點及不同點，它們各自碰到的局限性及未來發(fā)展的展望。最后，我們再來看模式識別在未來幾十年所面臨的挑戰(zhàn),這個挑戰(zhàn)主要是解決在學習和識別更大范圍適用性時所碰到的為建立相應處理的模型的脆弱問題。再有就是需要發(fā)展新的模式識別形式。介紹 對于人類的識別能力我們是非常熟悉的。因為我們在早些年就已經(jīng)會開發(fā)識別聲音、臉、動物、水果或簡單不動的東西的技術了。在開發(fā)出說話技術之前，一個象球的東西，甚至看上去只是象個球，就已經(jīng)

46、可以被識別出來了。所以除了記憶，抽象和推廣能力是推進模式識別技術的關鍵技術。最近幾年我們已可以處理更復雜的模式，這種模式可能不是直接基于通過感知器觀察出來的。 例如，我們能夠觀察發(fā)現(xiàn)某個討論會的中心議題或人與人之間關系的微妙的模式。后面一種模式是可能可以被明顯觀察到，例如傾聽某人在新的工作中因人際關系問題而產(chǎn)生的抱怨，我們不用切身其中就能夠發(fā)現(xiàn)這種相似和相同的例子，其復雜性莫過于人與人之間的社會相互影響。這里我們要學會區(qū)分只是從兩個例子中得到的模式。 模式識別的能力也可以在其它生物中被發(fā)現(xiàn)到：貓可以知道回家的路，狗能夠識別主人的腳印，蜜蜂會發(fā)現(xiàn)它要采蜜的花。這些例子中每一個直接聯(lián)結都是通過感觀

47、來實現(xiàn)的。不只是記憶方面，推廣能力是重要的一方面，從觀察到的相似事物中，雖然前后不一樣，也能夠進行識別，發(fā)現(xiàn)動物是怎么做到這一點是一個科學挑戰(zhàn)。 科學問題可以通過建立模型來解決，更確切的說是建立模擬器，例如人工系統(tǒng)通過學習來粗略地展示具有相同功能的東西，在建立這個系統(tǒng)和取得真實對象相關參數(shù)的過程中獲得得了對這個事物的理解，這樣的系統(tǒng)可以替換原來的對象，甚至可以提高原來的性能，但在其它方面可能是更差。例如，飛機可以飛得比鳥快，但在智能方面卻遠遠不如鳥，然而，我們的研究不是為了達到跟鳥全部一樣，更重要的是飛行能力。 通過模仿鳥的飛行可以學到很多飛行方面的技術，但無法學到其精確的分辨能力。通過建立固

48、定不動的翅膀，而不是自由扇動的翅膀，我們知道了怎么飛行。 最后，存在希望逐漸從原來的科學問題中引申出來的工程技術，如在重載下怎么飛得更長時間，怎么減少噪音，慢慢地把注意點轉移到其它的知識領域。 上面表明，模式識別(源于觀察的抽象和歸納能力)科學研究和應用技術領域中的人工智能模式識別設備設計存在差別，后者不會放過任何相互間互利的因素。注意這里所說的模式在很多層次上是有區(qū)分的，就如結構元素的簡單特征（如筆畫），體現(xiàn)了從在一組個體中表示某一個性質集的個體特征，到綜合概念和歸納的特征。一個模式可能表示成一個單獨個體，如某個總體、模型或概念的表示。結合模式、規(guī)律、特征或性質，模式識別所做的事可以說是在感

49、觀測定的低層次上(如圖像的象素)，也可以說是在推理和有意義概念的高層層次上(如圖像中的人臉)。這里，我們注重在科學研究方面，如模式識別的實現(xiàn)途徑是什么，特別是我們在建立人工識別設備需要具備什么技術。 已經(jīng)有些人在討論基于模擬和建模嘗試的模式識別科學了。在開始的六十年里，其中有個叫Sayre的人做了關于感知器、模式識別和分類的哲學研究，他斷言分類方法在某些程度上可以被成功實現(xiàn)，但或許也會失敗。根據(jù)問題的一些情況我們可以進行模擬識別。雖然我們不能完全識別某個個體，但是我們至少可以根據(jù)需要把對象分類出來。識別和分類的Sayre區(qū)分方法跟模式識別的兩個傳統(tǒng)的學習方法有關：無監(jiān)督學習和有監(jiān)督學習，這個兩個方法可以實現(xiàn)識別和分類方法，科學家利用這個自動化工具來發(fā)現(xiàn)自然界中的規(guī)律。 無監(jiān)督學習(也稱為試探性分析或聚類分析)：這個方法給研究者一種在觀察中自動表示未確定模式(規(guī)律)方法，通過這種方法模式種類被確定(檢驗)了下來，依此，根據(jù)Sayre觀點，一個模式就可以被被識別出來了。 有監(jiān)督學習：是這樣的一個自動系統(tǒng)，檢驗(確定)已被研究者通過一種表示方法定義好了的模式，這就是通過評估來實現(xiàn)的自動分類方法。 盡管Sayre已做了相關