Search

The continuous non-revisiting genetic algorithm (cNrGA) uses the entire search history and parameter-less adaptive mutation to significantly enhance search performance. Storing the entire search history is natural and costs little when the number of fitness evaluations is small or moderate. However, if the number of evaluations required is substantial, some memory management is desirable. In this paper, we propose two pruning mechanisms to keep the memory used constant. They are least recently used pruning and random pruning. The basic idea is to prune a unit of memory when the memory threshold is reached and some new search information is required to be stored, thus keeping the overall memory used constant. Meanwhile, both pruning strategies naturally form parameter-less adaptive mutation operators. A study is carried out to evaluate the impact on performance caused by loss of search history information. Experimental results show that (1) both strategies can maintain the performance of cNrGA, up to the empirical limit when 90 % of the search history is not recorded, (2) cNrGA and its variants with constant memory outperform the real-coded genetic algorithm and the standard particle swarm optimization. By pre-extracting all the current prune-able history information and storing them into a list, namely, to-prune-list, the overhead of both pruning strategies becomes small. This suggests that cNrGA can be extended to use in situations when the number of fitness evaluations is much larger than before with no significant effect on statistical performance. This widens the applicability of cNrGA to include more practical problems that require larger number of fitness evaluations before converging to the global optima.

Although ordering-based pruning algorithms possess relatively high efficiency, there remains room for further improvement. To this end, this paper describes the combination of a dynamic programming technique with the ensemble-pruning problem. We incorporate dynamic programming into the classical ordering-based ensemble-pruning algorithm with complementariness measure (ComEP), and, with the help of two auxiliary tables, propose a reasonably efficient dynamic form, which we refer to as ComDPEP. To examine the performance of the proposed algorithm, we conduct a series of simulations on four benchmark classification datasets. The experimental results demonstrate the significantly higher efficiency of ComDPEP over the classic ComEP algorithm. The proposed ComDPEP algorithm also outperforms two other state-of-the-art ordering-based ensemble-pruning algorithms, which use uncertainty weighted accuracy and reduce-error pruning, respectively, as their measures. It is noteworthy that, the effectiveness of ComDPEP is just the same with that of the classical ComEP algorithm.

In many applications, top-k query is an important operation to return a set of interesting points in a potentially huge data space. The existing algorithms, either maintaining too many candidates, or requiring assistant structures built on the specific attribute subset, or returning results with probabilistic guarantee, cannot process top-k query on massive data efficiently. This paper proposes a sorted-list-based TKAP algorithm, which utilizes some data structures of low space overhead, to efficiently compute top-k results on massive data. In round-robin retrieval on sorted lists, TKAP performs adaptive pruning operation and maintains the required candidates until the stop condition is satisfied. The adaptive pruning operation can be adjusted by the information obtained in round-robin retrieval to achieve a better pruning effect. The adaptive pruning rule is developed in this paper, along with its theoretical analysis. The extensive experimental results, conducted on synthetic and real-life data sets, show the significant advantage of TKAP over the existing algorithms.

Ensemble learning is a method of combining learners to obtain more reliable and accurate predictions in supervised and unsupervised learning. However, the ensemble sizes are sometimes unnecessarily large which leads to additional memory usage, computational overhead and decreased effectiveness. To overcome such side effects, pruning algorithms have been developed; since this is a combinatorial problem, finding the exact subset of ensembles is computationally infeasible. Different types of heuristic algorithms have developed to obtain an approximate solution but they lack a theoretical guarantee. Error Correcting Output Code (ECOC) is one of the well-known ensemble techniques for multiclass classification which combines the outputs of binary base learners to predict the classes for multiclass data. In this paper, we propose a novel approach for pruning the ECOC matrix by utilizing accuracy and diversity information simultaneously. All existing pruning methods need the size of the ensemble as a parameter, so the performance of the pruning methods depends on the size of the ensemble. Our unparametrized pruning method is novel as being independent of the size of ensemble. Experimental results show that our pruning method is mostly better than other existing approaches.

As is well known, the Greedy Ensemble Pruning (GEP) algorithm, also called the Directed Hill Climbing Ensemble Pruning (DHCEP) algorithm, possesses relatively good performance and high speed. However, because the algorithm only explores a relatively small subspace within the whole solution space, it often produces suboptimal solutions of the ensemble pruning problem. Aiming to address this drawback, in this work, we propose a novel Randomized GEP (RandomGEP) algorithm, also called the Randomized DHCEP (RandomDHCEP) algorithm, that effectively enlarges the search space of the classical DHCEP while maintaining the same level of time complexity with the help of a randomization technique. The randomization of the classical DHCEP algorithm achieves a good tradeoff between the effectiveness and efficiency of ensemble pruning. Besides, the RandomDHCEP algorithm naturally inherits the two intrinsic advantages that a randomized algorithm usually possesses. First, in most cases, its running time or space requirements are smaller than well-behaved deterministic ensemble pruning algorithms. Second, it is easy to comprehend and implement. Experimental results on three benchmark classification datasets verify the practicality and effectiveness of the RandomGEP algorithm.

Classification is an important data mining task that discovers hidden knowledge from the labeled datasets. Most approaches to pruning assume that all dataset are equally uniform and equally important, so they apply equal pruning to all the datasets. However, in real-world classification problems, all the datasets are not equal and considering equal pruning rate during pruning tends to generate a decision tree with large size and high misclassification rate. We approach the problem by first investigating the properties of each dataset and then deriving data-specific pruning value using expert knowledge which is used to design pruning techniques to prune decision trees close to perfection. An efficient pruning algorithm dubbed EKBP is proposed and is very general as we are free to use any learning algorithm as the base classifier. We have implemented our proposed solution and experimentally verified its effectiveness with forty real world benchmark dataset from UCI machine learning repository. In all these experiments, the proposed approach shows it can dramatically reduce the tree size while enhancing or retaining the level of accuracy.

Ensemble pruning deals with the reduction of base classifiers prior to combination in order to improve generalization and prediction efficiency. Existing ensemble pruning algorithms require much pruning time. This paper presents a fast pruning approach: pattern mining based ensemble pruning (PMEP). In this algorithm, the prediction results of all base classifiers are organized as a transaction database, and FP-Tree structure is used to compact the prediction results. Then a greedy pattern mining method is explored to find the ensemble of size k. After obtaining the ensembles of all possible sizes, the one with the best accuracy is outputted. Compared with Bagging, GASEN, and Forward Selection, experimental results show that PMEP achieves the best prediction accuracy and keeps the size of the final ensemble small, more importantly, its pruning time is much less than other ensemble pruning algorithms.

Decision trees are well-known and established models for classification and regression. In this paper, we focus on the estimation and the minimization of the misclassification rate of decision tree classifiers. We apply Lidstone's Law of Succession for the estimation of the class probabilities and error rates. In our work, we take into account not only the expected values of the error rate, which has been the norm in existing research, but also the corresponding reliability (measured by standard deviations) of the error rate. Based on this estimation, we propose an efficient pruning algorithm, called k-norm pruning, that has a clear theoretical interpretation, is easily implemented, and does not require a validation set. Our experiments show that our proposed pruning algorithm produces accurate trees quickly, and compares very favorably with two other well-known pruning algorithms, CCP of CART and EBP of C4.5.

A CSP search algorithm, like FC or MAC, explores a search tree during its run. Every node of the search tree can be associated with a CSP created by the refined domains of unassigned variables. If the algorithm detects that the CSP associated with a node is insoluble, the node becomes a dead-end. A strategy of pruning "by analogy" states that the current node of the search tree can be discarded if the CSP associated with it is "more constrained" than a CSP associated with some dead-end node. In this paper we present a method of pruning based on the above strategy. The information about the CSPs associated with dead-end nodes is kept in the structures called responsibility sets and kernels. We term the method that uses these structures for pruning RKP, which is abbreviation of Responsibility set, Kernel, Propagation. We combine the pruning method with algorithms FC and MAC. We call the resulting solvers FC-RKP and MAC-RKP, respectively. Experimental evaluation shows that MAC-RKP outperforms MAC-CBJ on random CSPs and on random graph coloring problems. The RKP-method also has theoretical interest. We show that under certain restrictions FC-RKP simulates FC-CBJ. It follows from the fact that intelligent backtracking implicitly uses the strategy of pruning "by analogy."

Classification is an important problem in data mining. Given a database of records, each with a class label, a classifier generates a concise and meaningful description for each class that can be used to classify subsequent records. A number of popular classifiers construct decision trees to generate class models. These classifiers first build a decision tree and then prune subtrees from the decision tree in a subsequent i>pruning phase to improve accuracy and prevent “overfitting”. Generating the decision tree in two distinct phases could result in a substantial amount of wasted effort since an entire subtree constructed in the first phase may later be pruned in the next phase. In this paper, we propose PUBLIC, an improved decision tree classifier that integrates the second “pruning” phase with the initial “building” phase. In PUBLIC, a node is not expanded during the building phase, if it is determined that it will be pruned during the subsequent pruning phase. In order to make this determination for a node, before it is expanded, PUBLIC computes a lower bound on the minimum cost subtree rooted at the node. This estimate is then used by PUBLIC to identify the nodes that are certain to be pruned, and for such nodes, not expend effort on splitting them. Experimental results with real-life as well as synthetic data sets demonstrate the effectiveness of PUBLIC's integrated approach which has the ability to deliver substantial performance improvements.

We propose depth-limited heuristic search as a general paradigm for real-time problem solving in a dynamic environment. When combined with iterative-deepening, it provides the ability to commit to an action almost instantaneously, but allows the quality of that decision to improve as long as time is available. Once a deadline is reached, the best decision arrived at is executed. We illustrate the paradigm in three different settings, corresponding to single-agent search, two-player games, and multi-agent problem solving. First we review two-player minimax search with alpha-beta pruning. Minimax can be extended to themaxn algorithm for more than two players, which admits a much weaker form of alpha-beta pruning. Finally, we explore real-time search algorithms for single-agent problems. Minimax is specialized tominimin, which allows a very powerfulalpha pruning algorithm. In addition,real-time-A * allows backtracking while still guaranteeing a solution and making locally optimal decisions.

This paper compares five methods for pruning decision trees, developed from sets of examples. When used with uncertain rather than deterministic data, decision-tree induction involves three main stages—creating a complete tree able to classify all the training examples, pruning this tree to give statistical reliability, and processing the pruned tree to improve understandability. This paper concerns the second stage—pruning. It presents empirical comparisons of the five methods across several domains. The results show that three methods—critical value, error complexity and reduced error—perform well, while the other two may cause problems. They also show that there is no significant interaction between the creation and pruning methods.