Abstract: Because user preferences change over time, capturing them is essential to predicting user behavior. Many shallow and deep recommendation algorithms model user static and dynamic preferences separately, making it difficult to fuse them for…
Abstract: Class imbalance problem that characterized with the skew distribution towards the majority arises as one challenge in recent years. Many oversampling techniques have been proposed to cope with this problem and some of them…
Abstract: Link prediction (LP) in networks predicts future interactions between elements and is a crucial machine-learning tool in genomics, social networks, marketing, and e-commerce recommender systems. Many LP techniques have been developed in the past,…
Abstract: Fuzzy discrete event systems (FDESs) can model event-driven systems as fuzzy automata with ambiguous states and event-invoked state transitions. In two recent papers, we developed algorithms for online-supervised learning of the fuzzy automaton’s event…
Abstract: Many successful methods have been proposed for learning low-dimensional representations on large-scale networks, but almost all are designed in inseparable processes, learning embeddings for entire networks even when only a small percentage of nodes…
Abstract: Complex factors make industrial power load forecasting difficult. Machine-learning methods are used to forecast industrial power load. Thus, a new power load forecasting method takes into account load characteristics in different regions, industries, and…
Abstract: Most head pose estimation methods require accurate ground-truth poses. The “ground truth” pose is often obtained subjectively by asking subjects to stare at different markers on the wall. Soft labels are better than explicit…
Abstract: Existing cross-domain recognition methods generally focus on changing data feature representation or classifier parameters, and their performance indicates their efficiency. Most methods do not integrate them into a single optimization objective to improve learning…
Abstract: Card fraud costs cardholders and issuers a lot. Modern methods detect fraud from transaction records using machine learning. Manually generating features requires domain knowledge and may be behind fraud’s modus operandi, so the online…
Abstract: Diverse ensembles are widely used in machine learning and credited with their success. Most of these methods increase ensemble diversity by sampling data or changing model structures. However, a family of ensemble models explicitly…
Abstract: Graph Neural Networks (GNNs) excel at recommendation, node classification, and link prediction. Graph neural network models create node embeddings by combining node features with neighboring node information. Most GNN models use a single aggregator…
Abstract: Machine learning’s fairness is a growing concern. Most machine-learning research is focused on either supervised or unsupervised learning. Two observations led us to consider semi-supervised learning for discrimination problems. First, previous research suggested that…
WhatsApp us