C.J. Thornton. Techniques of Computational Learning: an introduction. Chapman and Hall, 1992.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Assuming the original set consists of A 1 ; A 2 ; A n features, these variants can be defined below. Feature construction is a process that discovers missing information about the relationships between features and augments the space of features by inferring or creating additional features [5, 7, 6]. After feature construction, we may have additional m features A n 1 ; A n 2 ; A n m . For example, a new feature A k (n k n m) could be constructed by performing a logical operation on A i and A j from the original set. Another example is: a two dimensional problem (say, A 1 =width and A ....

C.J. Thornton. Techniques of Computational Learning: an introduction. Chapman and Hall, 1992.

....patterns. The distinction of this work from the related work mainly lies in how to efficiently add these conditions in a rule plus some special considerations for particular applications. Details will be given later. Another line of related work is of AQ11 and its variations [3] However, AQ11 [18] is a candiate elimnation algorithm that requires a generalization hierarchy in the search 2 of learned concepts. This is an advantage when the domain knowledge is available. Here we assume no domain knowledge. In the following section, we describe the base algorithm. In Section 3, we explain ....

C.J. Thornton. Techniques of Computational Learning: an introduction. Chapman and Hall, 1992.

....of improving the agent s ability to achieve its goals. I look at a definition of learning in information theoretic terms, as a many to few mapping in which the inputs are mapped to related outputs with maximal information preservation and introduction of the minimum of ambiguity (equivocation) Thornton (1992) looks at the definition of the similarity based learning paradigm in these terms; I apply similar techniques to the definition of reinforcement learning, in particular habituation and classical conditioning. A common element in definitions of attention is selection at one level or another; this ....

....input to target mapping) in order to reduce uncertainty about the environment by increasing the receiver s knowledge of the environment. However, in performing such a mapping there will be loss of information which can be seen as introduction of ambiguity,the information equivocation tradeoff (Thornton, 1992). Thus learning should also reduce ambiguity (also called entropy) by preserving information at the same time as performing this many to few mapping and thus minimising the ambiguity introduced. According to this definition, then, selection can be seen as the process or collection of processes by ....

Thornton, C. J. (1992). Techniques in Computational Learning: An Introduction. London: Chapman & Hall.

....Stone, 1966] ID3 has two new features that improved the algorithm. First an information theoretic splitting heuristic was used to enable small and efficient decision trees to be constructed. Second, the incorporation of windowing process that enabled the algorithm to cope with large training sets[Thornton, 1992]. With these advantages ID3 has become a mainstream of symbolic learning approaches and a number of derivatives are proposed by many researchers. For example, ID4 which incrementally builds a decision tree based on individually observed instances by maintaining positive and negative instance ....

....decision(class) attribute with two possible values, yes, no , n=2) In this example, Size , Color and Shape are attributes with sets of possible values small, medium, large , red, blue, green , and brick, wedge, sphere, pillar respectively. TABLE 1. Object Classification Training Set[Thornton, 1992]. Example no. Size Color Shape Decision 1 medium blue brick yes 2 small red wedge no 3 small red sphere yes 4 large red wedge no 5 large green pillar yes 6 large red pillar no 7 large green sphere yes Since n is two, the first step of the algorithm generates two sub tables which are shown in ....

Thornton, C.J. (1992). Techniques in Computational Learning-An Introduction, London: Chapman & Hall.

.... Trust and supervised by Professor Aaron Sloman I look at a definition of learning in information theoretic terms, as a many to few mapping in which the inputs are mapped to related outputs with maximal information preservation and introduction of the minimum of ambiguity (equivocation) Thornton (1992) looks at the definition of the similarity based learning paradigm in these terms; I apply similar techniques to the definition of reinforcement learning, in particular habituation and classical conditioning. A common element in definitions of attention is selection at one level or another; this ....

....input to target mapping) in order to reduce uncertainty about the environment by increasing the receiver s knowledge of the environment. However, in performing such a mapping there will be loss of information which can be seen as introduction of ambiguity, the information equivocation tradeoff (Thornton, 1992). Thus learning should also reduce ambiguity (also called entropy) by preserving information at the same time as performing this many to few mapping and thus minimising the ambiguity introduced. According to this definition, then, selection can be seen as the process or collection of processes by ....

Thornton, C. (1992). Techniques in Computational Learning: An Introduction. London: Chapman & Hall.

....[9] backpropagation relies primarily on the exploitation of statistical effects and is thus unable to deal properly with neutral mappings. There are several arguments in favour of this hypothesis. First, the backpropagation learning algorithm is a generalization of the leastmean squares algorithm [10] (and perceptron learning algorithm [11] which is effectively an iterative method for deriving statistical input output correlations. Thus the backpropagation learning method is based on a method for exploiting statistical effects. Second, the generalization performance observed in the 4 bit ....

Thornton, C. (1992). Techniques in Computational Learning: An Introduction. London: Chapman & Hall.

No context found.

Thornton, C.J., 1992. Techniques in Computational Learning-An Introduction. London: Chapman and Hall.

No context found.

Thornton, C. J. (1992b). Techniques in Computational Learning: An Introduction. Chapman and Hall, London.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC