Existing algorithms aiming to learn a binary classifier from positive (P) and unlabeled (U) data require estimating the class prior or label noise ahead of building a classification model. However, the estimation and classifier learning are normally conducted in a pipeline instead of being jointly optimized. In this paper, we propose to alternatively train the two steps using reinforcement learning. Our proposal adopts a policy network to adaptively make assumptions on the labels of unlabeled data, while a classifier is built upon the output of the policy network and provides rewards to learn a better policy. The dynamic and interactive training between the policy maker and the classifier can exploit the unlabeled data in a more effective manner and yield a significant improvement in terms of classification performance. Furthermore, we present two different approaches to represent the actions taken by the policy. The first approach considers continuous actions as soft labels, while the other uses discrete actions as hard assignment of labels for unlabeled examples. We validate the effectiveness of the proposed method on two public benchmark datasets as well as one ecommerce dataset. The results show that the proposed method is able to consistently outperform state-of-the-art methods in various settings.

Collaborative filtering (CF) is a technique used in recommender systems to provide meaningful suggestions based on known feedback obtained from like-minded users. The measure of similarity plays a critical role in the performance of neighborhood-based CF methods. However, conventional similarity measures suffer from limitations because they only consider the direction of the rating vectors. We propose a novel similarity measure that considers the semantic nuances of the ratings; in particular, it weights the contributions of ratings in proportion to the users' degree of indifference towards the items. Additionally, to address the sparsity problem that affects the performance of CF techniques, we propose a switching hybrid method that predicts user ratings based on either our custom similarity measure or through user and item biases. We evaluated the proposed method on six different datasets and compared it with other CF methods. The results show that the proposed recommender consistently outperforms those using conventional similarity measures when the sparsity of the dataset is high.

Representing products as a combination of properties that capture the essence of consumer sentiment is critical for companies that strive to understand consumer behavior. A catalogue of products described in terms of their attributes could offer companies a wide range of benefits; from improving existing products or developing new ones, to improving the quality of site search and offering better item recommendations to users. In this paper, we propose a method that encodes products as a sequence of attributes, each of which represents a different dimension of the consumer perception. In the proposed method, first, a base product set with known attribute values is built based on consumers' perceptions. Then, new product attribute vectors are estimated using product similarity. The proposed method also incorporates a new similarity measure that is based on purchase behavior and which is suitable for estimating product attribute vector distances. Because it takes into account the magnitude of the individual components of the vectors under comparison, the proposed method is free from the limitations of conventional similarity measures. The results of experiments conducted using real-world data indicate that the proposed method has superior performance compared to conventional approaches in terms of mean absolute error (MAE) and root mean squared error (RMSE).

No abstract

No abstract