Search

Agricultural data classification attracts more and more attention in the research area of intelligent agriculture. As a kind of important machine learning methods, ensemble learning uses multiple base classifiers to deal with classification problems. The rough set theory is a powerful mathematical approach to process unclear and uncertain data. In this paper, a rough set based ensemble learning algorithm is proposed to classify the agricultural data effectively and efficiently. An experimental comparison of different algorithms is conducted on four agricultural datasets. The results of experiment indicate that the proposed algorithm improves performance obviously.

Research highlights? A novel semi-supervised classification algorithm based on tolerance rough set and ensemble learning is proposed. ? Tolerance rough set theory is introduced to approximate concepts existed in documents and extract an initial set of negative example. ? Ensemble classifier is constructed to run iteratively for improving the approximation of negative data. Text classification has received more and more attention due to the enormous growth of digital content available on-line. This paper investigates the design of two-class text classifiers using positive and unlabeled data only. The specialty of this problem is that there is no labeled negative example for learning, which makes traditional text classification techniques inapplicable. In this paper, a novel semi-supervised classification algorithm based on tolerance rough set and ensemble learning is proposed. Tolerance rough set theory is used to approximate concepts existed in documents and extract an initial set of negative example. Then, SVM, Rocchio and Naive Bayes algorithms are used as base classifiers to construct an ensemble classifier, which runs iteratively and exploits margins between positive and negative data to progressively improve the approximation of negative data. Thus, the class boundary eventually converges to the true boundary of the positive class in the feature space. An experimental evaluation of different methods is carried out on two common text corpora, i.e., the Reuters-21578 collection and the WebKB collection. The experimental results indicate that the proposed method achieves significant performance improvement.

Automatic classification of text documents, one of essential techniques for Web mining, has always been a hot topic due to the explosive growth of digital documents available on-line. In text classification community, k-nearest neighbor (kNN) is a simple and yet effective classifier. However, as being a lazy learning method without premodelling, kNN has a high cost to classify new documents when training set is large. Rocchio algorithm is another well-known and widely used technique for text classification. One drawback of the Rocchio classifier is that it restricts the hypothesis space to the set of linear separable hyperplane regions. When the data does not fit its underlying assumption well, Rocchio classifier suffers. In this paper, a hybrid algorithm based on variable precision rough set is proposed to combine the strength of both kNN and Rocchio techniques and overcome their weaknesses. An experimental evaluation of different methods is carried out on two common text corpora, i.e., the Reuters-21578 collection and the 20-newsgroup collection. The experimental results indicate that the novel algorithm achieves significant performance improvement.

Signal prediction has attracted more and more attention from data mining and machine learning communities. Decision stump is a one-level decision tree, and it classifies instances by sorting them based on feature values. The boosting is a kind of powerful ensemble method and can improve the performance of prediction significantly. In this paper, boosting and decision stump algorithm are combined to analyse and predict the signal data. An experimental evaluation is carried out on the public signal dataset and the experimental results show that the boosting and decision stump-based algorithm clearly improves performance of signal prediction.

The agricultural data classification is a hot topic in the field of precision agriculture. Support vector machine (SVM) is a kind of structural risk minimization based learning algorithms. As a popular machine learning algorithm, SVM has been widely used in many fields such as information retrieval and text classification in the last decade. In this paper, SVM is introduced to classify the agricultural data. An experimental evaluation of different methods is carried out on the public agricultural dataset. Experimental results show that the SVM algorithm outperforms two popular algorithms, i.e., naive bayes and artificial neural network in terms of the F 1 measure.

Human Computer Interaction (HCI) challenges in mobile computing can be addressed by tailoring access and use of mobile services to user preferences. Our investigation of existent approaches to personalisation in context-aware computing found that user preferences are assumed to be static across different context descriptions, whilst in reality some user preferences are transient and vary with the change in context. Furthermore, existent preference models do not give an intuitive interpretation of a preference and lack user expressiveness. To tackle these issues, this paper presents a user preference model and mining framework for a context-aware m-services environment based on an intuitive quantitative preference measure and a strict partial order preference representation. Experimental evaluation of the user preference mining framework in a simulated m-Commerce environment showed that it is very promising. The preference mining algorithms were found to scale well with increases in the volumes of data.