Search

Automated algorithm design is attracting considerable recent research attention in solving complex combinatorial optimisation problems, due to that most metaheuristics may be particularly effective at certain problems or certain instances of the same problem but perform poorly at others. Within a general algorithm design framework, this study investigates reinforcement learning on the automated design of metaheuristic algorithms. Two groups of features, namely search-dependent and instance-dependent features, are firstly identified to represent the search space of algorithm design to support effective reinforcement learning on the new task of algorithm design. With these key features, a state-of-the-art reinforcement learning technique, namely proximal policy optimisation, is employed to automatically combine the basic algorithmic components within the general framework to develop effective metaheuristics. Patterns of the best designed algorithm, in particular the utilisation and transition of algorithmic components, are investigated. Experimental results on the capacitated vehicle routing problem with time windows benchmark dataset demonstrate the effectiveness of the identified features in assisting automated algorithm design with the proposed reinforcement learning model. • Two groups of features are identified to describe search space of algorithm design. • PPO is employed to extract useful knowledge hidden in the search data. • The influence of state representation (with different features) is investigated. • The utilisation and transition of algorithmic components are analysed. [ABSTRACT FROM AUTHOR]

• Calculate the similarity between pseudo-labels to keep complete sample information. • The latent representation space is learned in latent feature space and data space. • The latent representation contains the correlation of pseudo-labels. • Latent feature space provides discriminative information to guide feature selection. • Impose non-negative and l 2,1-2 -norm non-convex constraint on Q. In unsupervised feature selection, the relationship between pseudo-labels is often ignored, and the interconnection information between the data is not fully utilized. In order to solve these problems, this paper proposes a feature selection method via non-convex constraint and latent representation learning with Laplacian embedding (NLRL-LE). NLRL-LE keeps the correlation between the pseudo-labels to make the pseudo-label closer to the true label. And it combines with the interconnection information between data, learns the latent representation matrix to guide feature selection. Specifically, first, NLRL-LE regards each pseudo-label as a latent feature of the sample, constructs a latent feature graph, and retains the inherent attributes of the pseudo-labels. Second, latent representation learning is performed in the space which is made up of the latent feature space and data space. Since the latent feature graph retains the correlation between pseudo-labels, latent representation learning considers the interconnection information between data, and the information contained in the latent representation space is more complete. In addition, in order to make full use of pseudo-labels, the learned latent representation matrix is used as pseudo-label information to provide cluster labels in the latent representation space to guide feature selection. Finally, non-negative and l 2,1-2 -norm non-convex constraint are applied to the feature transformation matrix. The combination of non-negative constraint and non-convex constraint, compared with convex constraint, can ensure the row sparsity of the feature transformation matrix, select low-redundant features, and improve the feature selection effect. The experimental results show that the ACC and NMI of the NLRL-LE are better than the other seven compared algorithms on twelve datasets. [ABSTRACT FROM AUTHOR]

In recent years, geometric deep learning methods have been proposed, which are called Graph Convolutional Neural Networks (GCNNs). GCNNs not only can extract effective features like the classical CNN, but also can effectively reflect the true geometric structure of original data. Although GCNNs consider the geometric structure of original data, they construct the same feature graph to perform graph convolution, and ignore the difference between the local structures of different samples. Therefore, a novel Graph Convolutional Neural Network with Geometric and Discrimination information (GDGCNN) is proposed, which integrates traditional machine learning ideas to further improve the performance of feature extraction. In order to exploit differences between the local structures of different samples and make full use of the geometric structure of original data, GDGCNN constructs different feature graphs for different training batches to fully exploit the local geometry of data. Moreover, the discriminant regularization is introduced into GDGCNN to effectively utilize the discriminant information contained in original data. Therefore, GDGCNN has good discriminative ability and robustness. The experimental results show that GDGCNN can perform feature extraction tasks very well, and it is superior to some existing methods for classification in terms of accuracy and F1-Score. • GDGCNN constructs different feature graphs for different training batches to address the issue in GCNNs of neglecting the difference between the local structures of different samples in the dataset. Therefore, it can fully exploit the local geometry of original data. • GDGCNN effectively uses the discrimination information of original data to extract more discriminative features, and thus it has better learning ability and discriminating ability. • GDGCNN integrates traditional machine learning ideas into the new framework of GCNN to further improve the performance of feature extraction. [ABSTRACT FROM AUTHOR]