Knowledge graphs are useful for many AI tasks but often have missing facts. To populate the graphs, knowledge graph embedding models have been developed. TransE is one of such models and the first translation-based method. TransE is well known because the principle of TransE can effectively capture the rules of a knowledge graph although it seems very simple. However, TransE has problems with its regularization and unchangeable ratio of negative sampling. In this paper, we generalize TransE to solve these problem by proposing Knowledge Graph Embedding on a Lie Group (KGLG) and the Weighted Negative Part (WNP) method for the objective function of translation-based models. KGLG is the novel translation-based method which embeds entities and relations of a knowledge graph on any Lie group. It allows us not to employ regularization during training of the model if we choose a compact lie group for the embedding space. The WNP method is for changing the ratio of negative sampling, which enhances translation-based models. Our approach outperforms other state-of-the-art approaches such as TransE, DistMult and ComplEx on a standard link prediction task. We show that TorusE, KGLG on a torus, is scalable to large-size knowledge graphs and faster than the original TransE.

The Linked Open Data cloud contains tremendous amounts of interlinked instances with abundant knowledge for retrieval. However, because the ontologies are large and heterogeneous, it is time-consuming to learn all the ontologies manually and it is difficult to learn the properties important for describing instances of a specific class. To construct an ontology that helps users to easily access various data sets, we propose a semi-automatic system, called the Framework for InTegrating Ontologies, that can reduce the heterogeneity of the ontologies and retrieve frequently used core properties for each class. The framework consists of three main components: graph-based ontology integration, machine-learning-based approach for finding the core ontology classes and properties, and integrated ontology constructor. By analyzing the instances of linked data sets, this framework constructs a high-quality integrated ontology, which is easily understandable and effective in knowledge acquisition from various data sets using simple SPARQL queries.

Deep reinforcement learning (DRL) algorithms rely on carefully designed environment rewards that are extrinsic to the agent. However, in many real-world scenarios rewards are sparse or delayed, motivating the need for discovering efficient exploration strategies. While intrinsically motivated agents hold promise of better local exploration, solving problems that require coordinated decisions over long-time horizons remains an open problem. We postulate that to discover such strategies, a DRL agent should be able to combine local and high-level exploration behaviors. To this end, we introduce the concept of fast and slow curiosity that aims to incentivize long-time horizon exploration. Our method decomposes the curiosity bonus into a fast reward that deals with local exploration and a slow reward that encourages global exploration. We formulate this bonus as the error in an agent’s ability to reconstruct the observations given their contexts. We further propose to dynamically weight local and high-level strategies by measuring state diversity. We evaluate our method on a variety of benchmark environments, including Minigrid, Super Mario Bros, and Atari games. Experimental results show that our agent outperforms prior approaches in most tasks in terms of exploration efficiency and mean scores.

Reinforcement learning methods rely on rewards provided by the environment that are extrinsic to the agent. However, many real-world scenarios involve sparse or delayed rewards. In such cases, the agent can develop its own intrinsic reward function called curiosity to enable the agent to explore its environment in the quest of new skills. We propose a novel end-to-end curiosity mechanism for deep reinforcement learning methods, that allows an agent to gradually acquire new skills. Our method scales to high-dimensional problems, avoids the need of directly predicting the future, and, can perform in sequential decision scenarios. We formulate the curiosity as the ability of the agent to predict its own knowledge about the task. We base the prediction on the idea of skill learning to incentivize the discovery of new skills, and guide exploration towards promising solutions. To further improve data efficiency and generalization of the agent, we propose to learn a latent representation of the skills. We present a variety of sparse reward tasks in MiniGrid, MuJoCo, and Atari games. We compare the performance of an augmented agent that uses our curiosity reward to state-of-the-art learners. Experimental evaluation exhibits higher performance compared to reinforcement learning models that only learn by maximizing extrinsic rewards.

The increase of distributed energy resources in the smart grid calls for new ways to profitably exploit these resources, which can participate in day-ahead ancillary energy markets by providing flexibility. Higher profits are available for resource owners that are able to anticipate price peaks and hours of low prices or zero prices, as well as to control the resource in such a way that exploits the price fluctuations. Thus, this study presents a solution in which artificial neural networks are exploited to predict the day-ahead ancillary energy market prices. The study employs the frequency containment reserve for the normal operations market as a case study and presents the methodology utilized for the prediction of the case study ancillary market prices. The relevant data sources for predicting the market prices are identified, then the frequency containment reserve market prices are analyzed and compared with the spot market prices. In addition, the methodology describes the choices behind the definition of the model validation method and the performance evaluation coefficient utilized in the study. Moreover, the empirical processes for designing an artificial neural network model are presented. The performance of the artificial neural network model is evaluated in detail by means of several experiments, showing robustness and adaptiveness to the fast-changing price behaviors. Finally, the developed artificial neural network model is shown to have better performance than two state of the art models, support vector regression and ARIMA, respectively.