Automatic discovery of concepts has been an elusive area in machine learning. In this paper, we describe a system, called ADC, that automatically discovers concepts in a robotics domain, performing predicate invention...
详细信息
Automatic discovery of concepts has been an elusive area in machine learning. In this paper, we describe a system, called ADC, that automatically discovers concepts in a robotics domain, performing predicate invention. Unlike traditional approaches of concept discovery, our approach automatically finds and collects instances of potential relational concepts. An agent, using ADC, creates an incremental graph-based representation with the information it gathers while exploring its environment, from which common sub-graphs are identified. The subgraphs discovered are instances of potential relational concepts which are induced with inductive logic programming and predicate invention. Several concepts can be induced concurrently and the learned concepts can form arbitrarily hierarchies. The approach was tested for learning concepts of polygons, furniture, and floors of buildings with a simulated robot and compared with concepts suggested by users. (C) 2016 Elsevier B.V. All rights reserved.
Control flow compilation is a hybrid between classical WAM compilation and meta-call, limited to the compilation of non-recursive clause bodies. This approach is used successfully for the execution of dynamically gene...
详细信息
Control flow compilation is a hybrid between classical WAM compilation and meta-call, limited to the compilation of non-recursive clause bodies. This approach is used successfully for the execution of dynamically generated queries in an inductive logic programming setting (ILP). Control flow compilation reduces compilation times up to an order of magnitude, without slowing down execution. A lazy variant of control flow compilation is also presented. By compiling code by need, it removes the overhead of compiling unreached code (a frequent phenomenon in practical ILP settings), and thus reduces the size of the compiled code. Both dynamic compilation approaches have been implemented and were combined with query packs, an efficient ILP execution mechanism. It turns out that locality of data and code is important for performance. The experiments reported in the paper show that lazy control flow compilation is superior in both artificial and real life settings.
Background: Quantitative structure-activity relationships (QSAR) analysis of peptides is helpful for designing various types of drugs such as kinase inhibitor or antigen. Capturing various properties of peptides is es...
详细信息
Background: Quantitative structure-activity relationships (QSAR) analysis of peptides is helpful for designing various types of drugs such as kinase inhibitor or antigen. Capturing various properties of peptides is essential for analyzing two-dimensional QSAR. A descriptor of peptides is an important element for capturing properties. The atom pair holographic (APH) code is designed for the description of peptides and it represents peptides as the combination of thirty-six types of key atoms and their intermediate binding between two key atoms. Results: The substructure pair descriptor (SPAD) represents peptides as the combination of forty-nine types of key substructures and the sequence of amino acid residues between two substructures. The size of the key substructures is larger and the length of the sequence is longer than traditional descriptors. Similarity searches on C5a inhibitor data set and kinase inhibitor data set showed that order of inhibitors become three times higher by representing peptides with SPAD, respectively. Comparing scope of each descriptor shows that SPAD captures different properties from APH. Conclusion: QSAR/QSPR for peptides is helpful for designing various types of drugs such as kinase inhibitor and antigen. SPAD is a novel and powerful descriptor for various types of peptides. Accuracy of QSAR/QSPR becomes higher by describing peptides with SPAD.
To enable the treatment of hepatic metastasis with higher, theoretically more effective, doses of systemically toxic anticancer drugs, an isolated liver perfusion (ILP) technique was developed in WAG/Ola rats. First, ...
详细信息
To enable the treatment of hepatic metastasis with higher, theoretically more effective, doses of systemically toxic anticancer drugs, an isolated liver perfusion (ILP) technique was developed in WAG/Ola rats. First, in a toxicity study the maximally tolerated dose (MTD) of mitomycin C (MMC) was determined for a 25-min ILP and for hepatic artery infusion (HAI) after the administration of a bolus dose. The MTD in the ILP setting (4.8 mg/kg) was 4 times that using HAI (1.2 mg/kg). Subsequently, in a rat colorectal hepatic-metastasis model, concentrations of MMC in tumour, liver, plasma and perfusate were measured during a 25-min ILP to investigate the expected pharmacokinetic advantage of ILP. The mean plasma level determined after ILP (1.2 as well as 4.8 mg/kg MMC) was significantly lower (P < 0.001) than that obtained following HAI. This may explain both the absence of severe systemic toxicity and the higher MTD in ILP-treated groups. No significant difference in mean tumour and liver tissue concentrations of MMC were found when the groups treated with 1.2 mg/kg drug via HAI vs ILP were compared. The mean MMC concentration in tumour tissue was significantly higher (almost 5 times;P < 0.05) in rats treated by ILP with the MTD (4.8 mg/kg) than in those treated via HAI with the MTD (1.2 mg/kg). ILP of MMC can be safely performed using a dose 4 times higher than the MTD in the HAI setting, leading to an almost 5-fold concentration of MMC in hepatic metastasis. ILP of MMC may therefore represent a promising therapy for metastasis confined to the liver.
Central and peripheral insulin-like peptides (ILPs), which include insulin, insulin-like growth factor 1 (IGF1) and IGF2, exert many effects in the brain. Through their actions on brain growth and differentiation, ILP...
详细信息
Central and peripheral insulin-like peptides (ILPs), which include insulin, insulin-like growth factor 1 (IGF1) and IGF2, exert many effects in the brain. Through their actions on brain growth and differentiation, ILPs contribute to building circuitries that subserve metabolic and behavioural adaptation to internal and external cues of energy availability. In the adult brain each ILP has distinct effects, but together their actions ultimately regulate energy homeostasis - they affect nutrient sensing and regulate neuronal plasticity to modulate adaptive behaviours involved in food seeking, including high-level cognitive operations such as spatial memory. In essence, the multifaceted activity of ILPs in the brain may be viewed as a system organization involved in the control of energy allocation.
Dare2Del is an assistive system which facilitates intentional forgetting of irrelevant digital objects. For an assistive system to be helpful, the user has to trust the system's decisions. Explanations are a cruci...
详细信息
Dare2Del is an assistive system which facilitates intentional forgetting of irrelevant digital objects. For an assistive system to be helpful, the user has to trust the system's decisions. Explanations are a crucial component in establishing this trust. We will introduce different types of explanations which can vary along different dimensions such as level of detail and modality suitable for different application contexts. We will outline the cognitive companion system Dare2Del which is intended to support users managing digital objects in a working environment. Core of Dare2Del is an interpretable machine learning mechanism which induces decision rules to classify whether a digital objects is irrelevant. In this paper, we focus on irrelevance of files. We formalize the decision making process as logic inference. Finally, we present a method to generate verbal explanations for irrelevance decisions and point out how such explanations can be constructed on different levels of details. Furthermore, we show how verbal explanations can be related to the path context of the file. We conclude with a short discussion of the scope and restrictions of our approach.
We present our machine learning system, that uses inductive logic programming techniques to learn how to identify transmembrane domains from amino acid sequences. Our system facilitates the use of operators such as ...
详细信息
We present our machine learning system, that uses inductive logic programming techniques to learn how to identify transmembrane domains from amino acid sequences. Our system facilitates the use of operators such as 'contains', that act on entire sequences, rather than on individual elements of a sequence. The prediction accuracy of our new system is around 93%, and this compares favourably with earlier results.
We discuss how to learn non-recursive directed probabilistic logical models from relational data. This problem has been tackled before by upgrading the structure-search algorithm initially proposed for Bayesian networ...
详细信息
We discuss how to learn non-recursive directed probabilistic logical models from relational data. This problem has been tackled before by upgrading the structure-search algorithm initially proposed for Bayesian networks. In this paper we show how to upgrade another algorithm for learning Bayesian networks, namely ordering-search. For Bayesian networks, ordering-search was found to work better than structure-search. It is non-obvious that these results carry over to the relational case, however, since there ordering-search needs to be implemented quite differently. Hence, we perform an experimental comparison of these upgraded algorithms on four relational domains. We conclude that also in the relational case ordering-search is competitive with structure-search in terms of quality of the learned models, while ordering-search is significantly faster.
A particularly successful role for inductive logic programming (ILP) is as a tool for discovering useful relational features for subsequent use in a predictive model. Conceptually, the case for using ILP to construct ...
详细信息
A particularly successful role for inductive logic programming (ILP) is as a tool for discovering useful relational features for subsequent use in a predictive model. Conceptually, the case for using ILP to construct relational features rests on treating these features as functions, the automated discovery of which necessarily requires some form of first-order learning. Practically, there are now several reports in the literature that suggest that augmenting any existing feature with ILP-discovered relational features can substantially improve the predictive power of a model. While the approach is straightforward enough, much still needs to be done to scale it up to explore more fully the space of possible features that can be constructed by an ILP system. This is in principle, infinite and in practice, extremely large. Applications have been confined to heuristic or random selections from this space. In this paper, we address this computational difficulty by allowing features and models to be constructed in a distributed manner. That is, there is a network of computational units, each of which employs an ILP engine to construct some small number of features and then builds a (local) model. We then employ an asynchronous consensus-based algorithm, in which neighboring nodes share information and update local models. This gossip-based information exchange results in the formation of non-stationary Markov chains. For a category of models (those with convex loss functions), it can be shown (using the Supermartingale Convergence Theorem) that the algorithm will result in all nodes converging to a consensus model. In practice, it may be slow to achieve this convergence. Nevertheless, our results on synthetic and real datasets suggest that in relatively short time the "best" node in the network reaches a model whose predictive accuracy is comparable to that obtained using more computational effort in a non-distributed setting (the best node is identified as the one whose
A comparative study is presented of language biases employed in specific-to-general learning systems within the inductive logic programming (ILP) paradigm. More specifically, we focus on the biases employed in three w...
详细信息
A comparative study is presented of language biases employed in specific-to-general learning systems within the inductive logic programming (ILP) paradigm. More specifically, we focus on the biases employed in three well known systems: CLINT, GOLEM and ITOU, and evaluate both conceptually and empirically their strengths and weaknesses. The evaluation is carried out within the generic framework of the NINA system, in which bias is a parameter. Two different types of biases are considered: syntactic bias, which defines the set of well-formed clauses, and semantic bias, which imposes restrictions on the behaviour of hypotheses or clauses. NINA is also able to shift its bias (within a predefined series of biases), whenever its current bias is insufficient for finding complete and consistent concept definitions. Furthermore, a new formalism for specifying the syntactic bias of inductive logic programming systems is introduced.
暂无评论