Kleinberg, J., Papadimitriou, C. H. and Raghavan, P., "A microeconomic view of data mining" Data Mining and Knowledge Discovery, 2(4), 311--324, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....for pruning itemsets based on the support and utility constraints. In Section 5, we present some experimental results. 2 Related Work The necessity to develop methods for finding specific patterns which can be used to increase business utility has long been recognized by several researchers [7, 10, 14]. To the best of our knowledge, however, no work on association mining has been reported in the literature which formally models such patterns that are explicitly relating to a user s objective and its utility. Our work is related to but different from existing constrained association mining. ....

J. Kleinberg, C. Papadimitriou, and P. Raghavan. A microeconomic view of data mining. Journal of Data Mining and Knowledge Discovery, 6(1):83--105, 1998.

.... measures of interestingness have been put forward, such as unexpectedness [13] actionability [1] and rule templates [10] Finally, the most recent stream of research advocates the evaluation of the interestingness of associations in the light of the micro economic framework of the retailer [9]. More specifically, a pattern in the data is considered interesting only to the extent in which it can be used in the decision making process of the enterprise to increase its utility. It is in this latter stream of research that the authors have previously developed a model for product ....

J. Kleinberg, C. Papadimitriou, and P. Raghavan. A microeconomic view of data mining. In Knowledge Discovery and Data Mining, volume 2:4, pages 254--260. Kluwer Academic Publishers, 1998.

....applications remains rather limited [4, 5, 25] Nevertheless, the widespread acceptance of association rules as a valuable technique to solve real business problems will largely depend on the successful application of this technique on real world data. Moreover, it has been claimed recently [18] that the utility of extracted patterns (such as association rules) in decision making can only be addressed within the microeconomic framework of the enterprise. This means that a pattern in the data is interesting only to the extent in which it can be used in the decision making process of the ....

Kleinberg, J., Papadimitriou, C., and Raghavan, P. A microeconomic view of data mining. Data Mining and Knowledge Discovery Journal, 2 (4), 311-324, 1998.

....not give a clue on how to make such recommendations. Our work is similar in motivation to the actionability of patterns [ST96] the ability of the pattern to suggest concrete and profitable action by the decisionmakers. Recently, Kleinberg el at presented the microeconomic view of data mining [KPR98]. The microeconomic view approach is to maxxeDiecg(x, Yi) where g(x, Yi) is the utility of a decision x on a given customer i. In profit mining, we are to maxxeDg(x, C) where g is the total profit (a kind of utility) of a recommender x on uture customers, given the data about current customers ....

J. Kleinberg, C. Papadimitriou, and P. Raghavan. A microeconomic view of data mining. Journal of Knowledge Discovery and Data Mining, 1998, vol.2, 311-324 (also http://www.cs.berkeley. edu/christos/dml.ps)

....function: Q Gamma f j g k j=1 Delta = k X j=1 X x2 j x T c j ; 4) Intuitively, the objective function measures the combined coherence of all the k clusters. Such an objective function has also been proposed and studied theoretically in the context of market segmentation problems [Kleinberg et al. 1998]. 3.4 Spherical k means We seek a partitioning of the document vectors x 1 ; x 2 ; xn into k disjoint clusters 1 ; 2 ; k that maximizes the objective function in (4) that is, we seek a solution to the following maximization problem: f j g k j=1 = arg max f ....

.... ; 2 ; k that maximizes the objective function in (4) that is, we seek a solution to the following maximization problem: f j g k j=1 = arg max f j g k j=1 Q Gamma f j g k j=1 Delta : 5) Finding the optimal solution to the above maximization problem is NP complete [Kleinberg et al. 1998, Theorem 3.1] also, see [Garey et al. 1982] We now discuss an approximation algorithm, namely, the spherical k means algorithm, which is an effective and efficient iterative heuristic. 1. Start with an arbitrary partitioning of the document vectors, namely, f (0) j g k j=1 . Let fc (0) ....

Kleinberg, J., Papadimitriou, C. H., and Raghavan, P. (1998). A microeconomic view of data mining. Data Mining and Knowledge Discovery, 2(4):311--324.

....objective function: Q fp j g k j=1 = k j=1 x2p j x T c j ; 5) Intuitively, the objective function measures the combined coherence of all the k clusters. Such an objective function has also been proposed and studied theoretically in the context of market segmentation problems (Kleinberg et al. 1998). 3.4. SPHERICAL k MEANS We seek a partitioning of the document vectors x 1 ; x 2 ; x n into k disjoint clusters p 1 ; p 2 ; p k that maximizes the objective function in (5) that is, we seek a solution to the following maximization problem: fp j g k j=1 = ....

.... ; p 2 ; p k that maximizes the objective function in (5) that is, we seek a solution to the following maximization problem: fp j g k j=1 = arg max fp j g k j=1 Q fp j g k j=1 : 6) Finding the optimal solution to the above maximization problem is NP complete (Kleinberg et al. 1998, Theorem 3.1) also, see (Garey et al. 1982) We now discuss an approximation algorithm, namely, the spherical k means algorithm, which is an effective and efficient iterative heuristic. 1. Start with an arbitrary partitioning of the document vectors, namely, fp (0) j g k j=1 . Let fc ....

Kleinberg, J., C. H. Papadimitriou, and P. Raghavan: 1998, `A microeconomic view of data mining'. Data Mining and Knowledge Discovery 2(4), 311--324.

.... measures of interestingness have been put forward, such as unexpectedness [13] actionability [1] and rule templates [10] Finally, the most recent stream of research advocates the evaluation of the interestingness of associations in the light of the micro economic framework of the retailer [9]. More speci cally, a pattern in the data is considered interesting only to the extent in which it can be used in the decision making process of the enterprise to increase its utility. It is in this latter stream of research that the authors have previously developed a model for product selection ....

J. Kleinberg, C. Papadimitriou, and P. Raghavan. A microeconomic view of data mining. In Knowledge Discovery and Data Mining, volume 2:4, pages 254-260. Kluwer Academic Publishers, 1998.

....function: Q Gamma f j g k j=1 Delta = k X j=1 X x2 j x T c j ; 4) Intuitively, the objective function measures the combined coherence of all the k clusters. Such an objective function has also been proposed and studied theoretically in the context of market segmentation problems [Kleinberg et al. 1998]. 3.4 Spherical k means We seek a partitioning of the document vectors x 1 ; x 2 ; xn into k disjoint clusters 1 ; 2 ; k that maximizes the objective function in (4) that is, we seek a solution to the following maximization problem: f j g k j=1 = arg max f ....

.... ; 2 ; k that maximizes the objective function in (4) that is, we seek a solution to the following maximization problem: f j g k j=1 = arg max f j g k j=1 Q Gamma f j g k j=1 Delta : 5) Finding the optimal solution to the above maximization problem is NP complete [Kleinberg et al. 1998, Theorem 3.1] also, see [Garey et al. 1982] We now discuss an approximation algorithm, namely, the spherical k means algorithm, which is an effective and efficient iterative heuristic. 1. Start with an arbitrary partitioning of the document vectors, namely, f (0) j g k j=1 . Let fc (0) ....

Kleinberg, J., Papadimitriou, C. H., and Raghavan, P. (1998). A microeconomic view of data mining. Data Mining and Knowledge Discovery, 2(4):311--324.

....all such problems, it is NP hard, even in the unit weight case formulated above. One can define a segmentation problem (and in fact one of several variants) for every classical optimization problem. Segmentation problems are intended to capture certain aspects of the economic basis for data mining [11] and clustering; we explain this connection next. The Value of Data Mining Data mining is the application of statistical and machine learning techniques for extracting interesting patterns from raw data. Formalizing what interesting means in this context has been an important problem in the ....

....research in data mining deals with the efficient discovery of patterns for subsequent human evaluation of the degree to which they are interesting, and not on techniques for automatically evaluating mined patterns, or for automatically focusing on interesting patterns. We recently proposed in [11] a rigorous and algorithmic framework for such evaluation based on the pattern s utility in decision making. The framework formulated in [11] suggests a number of interesting computational issues, related to sensitivity analysis and clustering; in the present work we study algorithms for one class ....

[Article contains additional citation context not shown here]

J. Kleinberg, C. Papadimitriou, P. Raghavan, "A Microeconomic View of Data Mining," submitted to 1998 PODS.

No context found.

Kleinberg, J., Papadimitriou, C. H. and Raghavan, P., "A microeconomic view of data mining" Data Mining and Knowledge Discovery, 2(4), 311--324, 1998.

No context found.

J. Kleinberg, C. Papadimitriou, and P. Raghavan, "A microeconomic view of data mining," Knowl. Disc. Data Mining, vol. 2, no. 4, pp. 311--324, Dec. 1998.

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