Data Mining - Concepts and Techniques.

Ensemble methods are considered the state‐of‐the art solution for many machine learning challenges. Such methods improve the predictive performance of a single model by training multiple models and combining their predictions. This paper introduce the concept of ensemble learning, reviews traditional, novel and state‐of‐the‐art ensemble methods and discusses current challenges and trends in the field.

Ensemble learning: A survey

Introduction to stochastic programming

Feature selection is an important task in data mining and machine learning to reduce the dimensionality of the data and increase the performance of an algorithm, such as a classification algorithm. However, feature selection is a challenging task due mainly to the large search space. A variety of methods have been applied to solve feature selection problems, where evolutionary computation (EC) techniques have recently gained much attention and shown some success. However, there are no comprehensive guidelines on the strengths and weaknesses of alternative approaches. This leads to a disjointed and fragmented field with ultimately lost opportunities for improving performance and successful applications. This paper presents a comprehensive survey of the state-of-the-art work on EC for feature selection, which identifies the contributions of these different algorithms. In addition, current issues and challenges are also discussed to identify promising areas for future research.

/pdf/a-survey-on-evolutionary-computation-approaches-to-feature-5c2qznm5vr.pdf

A Survey on Evolutionary Computation Approaches to Feature Selection

C4.5: Programs for Machine Learning (書評)

The C4.5 decision tree (DT) can be applied in various fields and discovers knowledge for human understanding. However, different problems typically require different parameter settings. Rule of thumb or trial-and-error methods are generally utilized to determine parameter settings. However, these methods may result in poor parameter settings and unsatisfactory results. On the other hand, although a dataset can contain numerous features, not all features are beneficial for classification in C4.5 algorithm. Therefore, a novel scatter search-based approach (SS + DT) is proposed to acquire optimal parameter settings and to select the beneficial subset of features that result in better classification results. To evaluate the efficiency of the proposed SS + DT approach, datasets in the UCI (University of California, Irvine) Machine Learning Repository are utilized to assess the performance of the proposed approach. Experimental results demonstrate that the parameter settings for the C4.5 algorithm obtained by the SS + DT approach are better than those obtained by other approaches. When feature selection is considered, classification accuracy rates on most datasets are increased. Therefore, the proposed approach can be utilized to identify effectively the best parameter settings for C4.5 algorithm and useful features.

Parameter determination and feature selection for C4.5 algorithm using scatter search approach

The museum visitor routing problem

Through the use of an effective simulated annealing algorithm, this paper presents an extensive computational investigation concerning the performance evaluation of non-permutation vs. permutation schedules for the flowline manufacturing cell with sequence dependent setup times. The results show that there are significant improvements made by using non-permutation schedules over permutation schedules regarding completion-time based and due-date based criteria. It is suggested that practitioners could adopt non-permutation schedules to improve their operational efficiency within a reasonable amount of computational effort.

Permutation and non-permutation schedules for the flowline manufacturing cell with sequence dependent family setups

This paper studies the distributed blocking flow shop scheduling problem (DBFSP) using meta-heuristics. A mixed integer programming model for solving the problem is proposed, and then three versions of the hybrid iterated greedy algorithm (HIG1, HIG2, and HIG3) are developed, combining the advantages of an iterated greedy algorithm with the operators of the variable Tabu list, the constant Tabu list, and the cooling schedule. A benchmark problem set is used to assess empirically the performance of the HIG1, HIG2, and HIG3 algorithms. Computational results show that all the three versions of the proposed algorithm can efficiently and effectively minimize the maximum completion time among all factories of the DBFSP, and HIG1 is the most effective.

Minimizing Makespan in Distributed Blocking Flowshops Using Hybrid Iterated Greedy Algorithms

Data-mining algorithms have been used in many classification problems. Among them, the decision tree (DT), back-propagation network (BPN), and support vector machine (SVM) are popular and can be applied to various areas. Nevertheless, different problems may require different parameter values when applying DT, BPN or SVM. If parameter values are not set well, results may turn out to be unsatisfactory. Further, a dataset may contain many features; however, not all features are beneficial for classifications. Therefore, a scatter search (SS) approach is proposed to obtain the better parameters and select the beneficial subset of features to attain better classification results. The above classification algorithms have their respective advantages and disadvantages, and suitability is influenced by the characteristics of the problem. If the algorithms can function together in a so-called ensemble, it is expected that better results can be obtained. Therefore, this study adapts ensemble to further enhance the classification accuracy rate. In order to evaluate the performance of the proposed approach, datasets in UCI (University of California, Irvine) were applied as the test problem set. The corresponding results were compared to several well-known, published approaches. The comparative study shows that the proposed approach improved the classification accuracy rate in most datasets. Thus, the proposed approach can be useful to both practitioners and researchers.

Shih-Wei Lin

Papers

Parameter determination and feature selection for C4.5 algorithm using scatter search approach

The museum visitor routing problem

Permutation and non-permutation schedules for the flowline manufacturing cell with sequence dependent family setups

Minimizing Makespan in Distributed Blocking Flowshops Using Hybrid Iterated Greedy Algorithms

Enhancing the classification accuracy by scatter-search-based ensemble approach