H. Liu and R. Sedtiono. Incremental feature selection. Applied Intelligence, 9(3):217--230, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....applications, calculating an amount of accuracy estimations (traning and testing on significant amount of data) not envisioned in the 1980s. Before applying the wrapper approach, an enumeration of the available resources is quite critical; two main factors can make the selection problem large [33]: the number of features and the number of instances. One must bear in mind that in the wrapper approach, every possible solution visited by the search engine requires the time needed by the learning algorithm in the training phase. Many approaches have been porposed in order to alleviate the ....

H. Liu and R. Sedtiono. Incremental feature selection. Applied Intelligence, 9(3):217--230, 1998.

....process consists of the following stages: 1. A genetic feature selection process, that gets a feature subset to learn the FRBCS from it. The proposed feature selection process uses a GA as search algorithm and it has wrapper nature [35] We also use a feature selection algorithm with lter nature [38, 37, 39] that searches for a variable cardinality feature subset to obtain the chromosome length for our proposal of genetic feature selection process. We will explain this process in detail in the next section. 2. An iterative fuzzy rule generation process, which using only the selected features ....

....of feature selection algorithms in two steps: 1. The rst step looks for a feature subset with variable size considering class separability measures to determine an optimal feature number for a speci c classi cation problem. In this study, we employ the lter algorithm Las Vegas Filter (LVF) [38, 37, 39], which is based on the inconsistency rate, proposed by Liu and Setiono (which is described in Appendix A) This lter feature selection method provides us a cardinality for the feature subset with minimum number of inconsistencies. Besides this cardinality, we use the feature subset size given ....

[Article contains additional citation context not shown here]

H. Liu and R. Setiono. Incremental feature selection. Applied Intelligence, 9:217-230, 1998.

....an amount of accuracy estimations (training and testing on significants amounts of data) not envisioned in the 80 s. Before applying the wrapper approach, an enumeration of the available computer resources is critical. Two different factors become a FSS problem large (Liu and Setiono [57]) the number of features and the number of instances. One must bear in mind the time needed for the learning algorithm used in the wrapper scheme as a training phase is required for every possible solution visited by the FSS search engine. Many approaches have been proposed in literature to ....

.... induction algorithm in the evaluation function generate new FSS algorithms, such as FOCUS (Almuallin and Dietterich [4] RELIEF (Kira and Rendell [40] Cardie s algorithm [16] Koller and Sahami s work with probabilistic concepts [48] or the named Incremental Feature Selection (Liu and Setiono [57]) Nowadays, the filter approach is receiving considerable attention from the Data Mining community to deal with huge databases when the wrapper approach is unfeasible (Liu and Motoda [56] Figure 2 locates the role of filter and wrapper approaches within the overall FSS process. When the size ....

H. Liu, R. Setiono, Incremental Feature Selection, Applied Intelligence 9 (3) (1998) 217-230.

....design parameters and agents. This allows LEAD to relate agent actions (e.g. decision, critique, request, conflict etc. with the variation of a specific design parameter. 2) The covariational analysis component uses wrappers for relevant condition selection. Wrappers (Kohavi and John 1998; Liu and Setiono 1998) apply an induction algorithm to a training data set. The experiments are run by eliminating different sets of features from the training data instances. Specifically, wrappers eliminate conditions from the candidate condition set. The wrapper method proposes a subset of features that are ....

Liu, H., and R. Setiono (1998). Incremental Feature Selection. Applied Intelligence.

....of the resulting concept description are the relevant conditions that the agent has identified as influencing the occurrence of specific assertion ranges, i.e. classes. To achieve this learning goal, agents use wrappers for relevant condition selection (figure 6) Wrappers (Kohavi and John 1998; Liu and Setiono 1998) apply an induction algorithm to a training data set. The experiments are run by eliminating different sets of features from the training data instances. Specifically, wrappers eliminate conditions from the candidate condition set. The wrapper method proposes a subset of features that are relevant ....

Liu, H., and R. Setiono (1998). Incremental Feature Selection. Applied Intelligence.

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