Publications
273 results found
Muggleton S, Dai WZ, Sammut C, et al., 2018, Meta-Interpretive Learning from noisy images, Machine Learning, Vol: 107, Pages: 1097-1118, ISSN: 0885-6125
Statistical machine learning is widely used in image classification. However, most techniques (1) require many images to achieve high accuracy and (2) do not provide support for reasoning below the level of classification, and so are unable to support secondary reasoning, such as the existence and position of light sources and other objects outside the image. This paper describes an Inductive Logic Programming approach called Logical Vision which overcomes some of these limitations. LV uses Meta-Interpretive Learning (MIL) combined with low-level extraction of high-contrast points sampled from the image to learn recursive logic programs describing the image. In published work LV was demonstrated capable of high-accuracy prediction of classes such as regular polygon from small numbers of images where Support Vector Machines and Convolutional Neural Networks gave near random predictions in some cases. LV has so far only been applied to noise-free, artificially generated images. This paper extends LV by (a) addressing classification noise using a new noise-telerant version of the MIL system Metagol, (b) addressing attribute noise using primitive-level statistical estimators to identify sub-objects in real images, (c) using a wider class of background models representing classical 2D shapes such as circles and ellipses, (d) providing richer learnable background knowledge in the form of a simple but generic recursive theory of light reflection. In our experiments we consider noisy images in both natural science settings and in a RoboCup competition setting. The natural science settings involve identification of the position of the light source in telescopic and microscopic images, while the RoboCup setting involves identification of the position of the ball. Our results indicate that with real images the new noise-robust version of LV using a single example (i.e. one-shot LV) converges to an accuracy at least comparable to a thirty-shot statistical machine learner on bot
Muggleton SH, Schmid U, Zeller C, et al., 2018, Ultra-strong machine learning: comprehensibility of programs learned with ILP, Machine Learning, Vol: 107, Pages: 1119-1140, ISSN: 0885-6125
During the 1980s Michie defined Machine Learning in terms of two orthogonal axes of performance: predictive accuracy and comprehensibility of generated hypotheses. Since predictive accuracy was readily measurable and comprehensibility not so, later definitions in the 1990s, such as Mitchell’s, tended to use a one-dimensional approach to Machine Learning based solely on predictive accuracy, ultimately favouring statistical over symbolic Machine Learning approaches. In this paper we provide a definition of comprehensibility of hypotheses which can be estimated using human participant trials. We present two sets of experiments testing human comprehensibility of logic programs. In the first experiment we test human comprehensibility with and without predicate invention. Results indicate comprehensibility is affected not only by the complexity of the presented program but also by the existence of anonymous predicate symbols. In the second experiment we directly test whether any state-of-the-art ILP systems are ultra-strong learners in Michie’s sense, and select the Metagol system for use in humans trials. Results show participants were not able to learn the relational concept on their own from a set of examples but they were able to apply the relational definition provided by the ILP system correctly. This implies the existence of a class of relational concepts which are hard to acquire for humans, though easy to understand given an abstract explanation. We believe improved understanding of this class could have potential relevance to contexts involving human learning, teaching and verbal interaction.
Cropper A, Muggleton SH, 2018, Learning efficient logic programs, Machine Learning, Pages: 1-21, ISSN: 0885-6125
© 2018 The Author(s) When machine learning programs from data, we ideally want to learn efficient rather than inefficient programs. However, existing inductive logic programming (ILP) techniques cannot distinguish between the efficiencies of programs, such as permutation sort (n!) and merge sort (Formula presented.). To address this limitation, we introduce Metaopt, an ILP system which iteratively learns lower cost logic programs, each time further restricting the hypothesis space. We prove that given sufficiently large numbers of examples, Metaopt converges on minimal cost programs, and our experiments show that in practice only small numbers of examples are needed. To learn minimal time-complexity programs, including non-deterministic programs, we introduce a cost function called tree cost which measures the size of the SLD-tree searched when a program is given a goal. Our experiments on programming puzzles, robot strategies, and real-world string transformation problems show that Metaopt learns minimal cost programs. To our knowledge, Metaopt is the first machine learning approach that, given sufficient numbers of training examples, is guaranteed to learn minimal cost logic programs, including minimal time-complexity programs.
Hocquette C, Muggleton S, 2018, How Much Can Experimental Cost Be Reduced in Active Learning of Agent Strategies?, Pages: 38-53, ISSN: 0302-9743
In science, experiments are empirical observations allowing for the arbitration of competing hypotheses and knowledge acquisition. For a scientist that aims at learning an agent strategy, performing experiments involves costs. To that extent, the efficiency of a learning process relies on the number of experiments performed. We study in this article how the cost of experimentation can be reduced with active learning to learn efficient agent strategies. We consider an extension of the meta-interpretive learning framework that allocates a Bayesian posterior distribution over the hypothesis space. At each iteration, the learner queries the label of the instance with maximum entropy. This produces the maximal discriminative over the remaining competing hypotheses, and thus achieves the highest shrinkage of the version space. We study the theoretical framework and evaluate the gain on the cost of experimentation for the task of learning regular grammars and agent strategies: our results demonstrate the number of experiments to perform to reach an arbitrary accuracy level can at least be halved.
Conn H, Muggleton SH, 2018, The effect of predicate order on curriculum learning in ILP, 27th International Conference on Inductive Logic Programming, Pages: 17-21, ISSN: 1613-0073
© by the paper's authors. Development of effective methods for learning large programs is arguably one of the hardest unsolved problems within ILP. The most obvious approach involves learning a sequence of predicate definitions incrementally. This approach is known as Curriculum Learning. However, Quinlan and Cameron-Jones' paper from 1993 indicates difficulties in this approach since the predictive accuracy of ILP systems, such as FOIL, rapidly degrades given a growing set of learned background predicates, even when a reasonable ordering over the predicate sequence is chosen. Limited progress was made in this problem until the recent advent of bias-reformulation methods within Meta-Interpretive Learning. In this paper we show empirically that given a well-ordered predicate sequence, relatively large sets of dyadic predicates can be learned incrementally using a state-of-the-art Meta-Interpretive Learning system which employs a Universal Set of metarules. However, further experiments show how progressive random permutations of the sequence rapidly degrades performance in a fashion comparable to Quinlan and Cameron-Jones's results. On the basis of these results we propose the need for further identification of methods for identifying well-ordered predicate sequences to address this issue.
Dai W-Z, Muggleton S, Wen J, et al., 2018, Logical Vision: One-Shot Meta-Interpretive Learning from Real Images, 27th International Conference on Inductive Logic Programming (ILP), Publisher: SPRINGER INTERNATIONAL PUBLISHING AG, Pages: 46-62, ISSN: 0302-9743
Schmid U, Muggleton SH, Singh R, 2017, Approaches and Applications of Inductive Programming (Dagstuhl Seminar 17382)., Dagstuhl Seminar 17382, Pages: 86-108
Tamaddoni-Nezhad A, Besold T, Schmid U, et al., 2017, How does Predicate Invention affect Human Comprehensibility?, Inductive Logic Programming, 26th International Conference
Schmid U, Zeller C, Besold T, et al., 2017, How does predicate invention affect human comprehensibility?, Pages: 52-67, ISSN: 0302-9743
During the 1980s Michie defined Machine Learning in terms of two orthogonal axes of performance: predictive accuracy and comprehensibility of generated hypotheses. Since predictive accuracy was readily measurable and comprehensibility not so, later definitions in the 1990s, such as that of Mitchell, tended to use a one-dimensional approach to Machine Learning based solely on predictive accuracy, ultimately favouring statistical over symbolic Machine Learning approaches. In this paper we provide a definition of comprehensibility of hypotheses which can be estimated using human participant trials. We present the results of experiments testing human comprehensibility of logic programs learned with and without predicate invention. Results indicate that comprehensibility is affected not only by the complexity of the presented program but also by the existence of anonymous predicate symbols.
Muggleton SH, 2017, Meta-Interpretive Learning: Achievements and Challenges, 1st International Joint Conference on Rules and Reasoning (RuleML+RR), Publisher: SPRINGER INTERNATIONAL PUBLISHING AG, Pages: 1-6, ISSN: 0302-9743
- Author Web Link
- Cite
- Citations: 1
Cropper A, Tamaddoni-Nezhad A, Muggleton SH, 2016, Meta-interpretive learning of data transformation programs, 25th International Conference, ILP 2015, Publisher: Springer, Pages: 46-59, ISSN: 0302-9743
Data transformation involves the manual construction of large numbers of special-purpose programs. Although typically small, such programs can be complex, involving problem decomposition, recursion, and recognition of context. Building such programs is common in commercial and academic data analytic projects and can be labour intensive and expensive, making it a suitable candidate for machine learning. In this paper, we use the meta-interpretive learning framework (MIL) to learn recursive data transformation programs from small numbers of examples. MIL is well suited to this task because it supports problem decomposition through predicate invention, learning recursive programs, learning from few examples, and learning from only positive examples. We apply Metagol, a MIL implementation, to both semi-structured and unstructured data. We conduct experiments on three real-world datasets: medical patient records, XML mondial records, and natural language taken from ecological papers. The experimental results suggest that high levels of predictive accuracy can be achieved in these tasks from small numbers of training examples, especially when learning with recursion.
Gill RJ, Woodward G, 2016, Networking our way to better ecosystem service provision, Trends in Ecology & Evolution, Vol: 31, Pages: 105-115, ISSN: 0169-5347
The ecosystem services (EcoS) concept is being used increasingly to attach values to natural systems and the multiple benefits they provide to human societies. Ecosystem processes or functions only become EcoS if they are shown to have social and/or economic value. This should assure an explicit connection between the natural and social sciences, but EcoS approaches have been criticized for retaining little natural science. Preserving the natural, ecological science context within EcoS research is challenging because the multiple disciplines involved have very different traditions and vocabularies (common-language challenge) and span many organizational levels and temporal and spatial scales (scale challenge) that define the relevant interacting entities (interaction challenge). We propose a network-based approach to transcend these discipline challenges and place the natural science context at the heart of EcoS research.
Cropper A, Muggleton SH, 2016, Learning higher-order logic programs through abstraction and invention, Pages: 1418-1424, ISSN: 1045-0823
Many tasks in AI require the design of complex programs and representations, whether for programming robots, designing game-playing programs, or conducting textual or visual transformations. This paper explores a novel inductive logic programming approach to learn such programs from examples. To reduce the complexity of the learned programs, and thus the search for such a program, we introduce higher-order operations involving an alternation of Abstraction and Invention. Abstractions are described using logic program definitions containing higher-order predicate variables. Inventions involve the construction of definitions for the predicate variables used in the Abstractions. The use of Abstractions extends the Meta-Interpretive Learning framework and is supported by the use of a userextendable set of higher-order operators, such as map, until, and ifthenelse. Using these operators reduces the textual complexity required to express target classes of programs. We provide sample complexity results which indicate that the approach leads to reductions in the numbers of examples required to reach high predictive accuracy, as well as significant reductions in overall learning time. Our experiments demonstrate increased accuracy and reduced learning times in all cases. We believe that this paper is the first in the literature to demonstrate the efficiency and accuracy advantages involved in the use of higher-order abstractions.
Gulwan S, Hernández-Orall J, Kitzelmann E, et al., 2015, Inductive Programming Meets the Real World, Communications of the ACM, Vol: 58, Pages: 90-99, ISSN: 0001-0782
Much of the world's population use computers for everyday tasks, but most fail to benefit from the power of computation due to their inability to program. Most crucially, users often have to perform repetitive actions manually because they are not able to use the macro languages available for many application programs. Recently, a first mass-market product was presented in the form of the Flash Fill feature in Microsoft Excel 2013. Flash Fill allows end users to automatically generate string-processing programs for spreadsheets from one or more user-provided examples. Flash Fill is able to learn a large variety of quite complex programs from only a few examples because of incorporation of inductive programming methods.
Muggleton SH, 2015, Learning efficient logical robot strategies involving composable objects, International Joint Conference Artificial Intelligence (IJCAI 2015), Pages: 3423-3429
Muggleton SH, Lin D, Tamaddoni-Nezhad A, 2015, Meta-interpretive learning of higher-order dyadic datalog: predicate invention revisited, MACHINE LEARNING, Vol: 100, Pages: 49-73, ISSN: 0885-6125
- Author Web Link
- Open Access Link
- Cite
- Citations: 87
Reynolds CR, Muggleton SH, Sternberg MJE, 2015, Incorporating virtual reactions into a logic-based ligand-based virtual screening method to discover new leads, Molecular Informatics, Vol: 34, Pages: 615-625, ISSN: 1868-1751
The use of virtual screening has become increasingly central to the drug development pipeline, with ligand-based virtual screening used to screen databases of compounds to predict their bioactivity against a target. These databases can only represent a small fraction of chemical space, and this paper describes a method of exploring synthetic space by applying virtual reactions to promising compounds within a database, and generating focussed libraries of predicted derivatives. A ligand-based virtual screening tool Investigational Novel Drug Discovery by Example (INDDEx) is used as the basis for a system of virtual reactions. The use of virtual reactions is estimated to open up a potential space of 1.21×1012 potential molecules. A de novo design algorithm known as Partial Logical-Rule Reactant Selection (PLoRRS) is introduced and incorporated into the INDDEx methodology. PLoRRS uses logical rules from the INDDEx model to select reactants for the de novo generation of potentially active products. The PLoRRS method is found to increase significantly the likelihood of retrieving molecules similar to known actives with a p-value of 0.016. Case studies demonstrate that the virtual reactions produce molecules highly similar to known actives, including known blockbuster drugs.
Dai WZ, Muggleton SH, Zhou ZH, 2015, Logical vision: Meta-interpretive learning for simple geometrical concepts, Pages: 1-16, ISSN: 1613-0073
Progress in statistical learning in recent years has enabled computers to recognize objects with near-human ability. However, recent studies have revealed particular drawbacks in current computer vision systems which suggest there exist considerable differences between the way these systems function compared with human visual cognition. Major differences are that: 1) current computer vision systems learn high-level notions directly from the low-level feature space and ignore the mid-level representations, which makes them difficult to incorporate background knowledge. 2) typical computer vision systems learn visual concepts discriminatively instead of encoding the knowledge necessary to produce a visual representation of the class. In this paper, we introduce a framework referred as Logical Vision which is demonstrated on learning visual concepts constructively and symbolically. Given a set of images, a set of first-order logic formulae of background knowledge and a set of examples of target visual concepts, Logical Vision extracts logical facts concerning geometrical elements from an image by sampling low-level features guided by the background knowledge and conjecturing geometrical elements as output. It first extracts logical facts of mid-level features, then generative Meta-Interpretive Learning technique is applied to learn high-level notions because it is capable of learning recursions, inventing predicates and so on. Owing to its symbolic representation paradigm, in our implementation, Logical Vision is fully implemented in Prolog apart from low-level image feature extraction primitives. In our implementation, Logical Vision was used to extract polygon edges as mid-level symbols, and a generalized Meta-Interpreter Learner was applied to learn high-level geometrical notions. Experiments are conducted on learning shapes (e.g. triangles, quadrilaterals, etc.), regular polygons and right-angle triangles. These demonstrates that learning visual concepts constructi
Farquhar C, Grov G, Cropper A, et al., 2015, Typed meta-interpretive learning for proof strategies, Pages: 17-32, ISSN: 1613-0073
Formal verification of computer programs is increasingly used in industry. A popular technique is interactive theorem proving, used for instance by Intel in HOL light. The ability to learn and re-apply proof strategies from a small set of proofs would significantly increase the productivity of these systems, and make them more cost-effective to use. Previous learning attempts have had limited success, which we believe is a result of missing key goal properties in the strategies. Capturing such properties requires predicate invention, and the only state-of-the-art ILP technique which supports this is meta-interpretive learning (MIL). We show that MIL is applicable to this problem, but that without type information it offers limited improvements in quality over previous work. We then extend MIL with types and give preliminary results indicating that this extension learns better-quality strategies with suitable goal properties. We also show that the quality of the learned strategies can be further enhanced through the use of dependent learning.
Hernández-Orallo J, Muggleton SH, Schmid U, et al., 2015, Approaches and Applications of Inductive Programming (Dagstuhl Seminar 15442)., Dagstuhl Reports, Vol: 5, Pages: 89-111
Tamaddoni-Nezhad A, Bohan D, Raybould A, et al., 2015, Towards Machine Learning of Predictive Models from Ecological Data, 24th International Conference on Inductive Logic Programming (ILP), Publisher: SPRINGER-VERLAG BERLIN, Pages: 154-167, ISSN: 0302-9743
- Author Web Link
- Cite
- Citations: 4
Cropper A, Muggleton SH, 2015, Logical Minimisation of Meta-Rules Within Meta-Interpretive Learning, 24th International Conference on Inductive Logic Programming (ILP), Publisher: SPRINGER-VERLAG BERLIN, Pages: 62-75, ISSN: 0302-9743
- Author Web Link
- Cite
- Citations: 11
Cropper A, Muggleton SH, 2015, Can predicate invention compensate for incomplete background knowledge?, Publisher: IOS Press, Pages: 27-36
Tamaddoni-Nezhad A, Lin D, Watanabe H, et al., 2014, Machine Learning of Biological Networks Using Abductive ILP, Logical Modeling of Biological Systems, Pages: 363-401, ISBN: 9781848216808
This chapter demonstrates the potential of a logic-based approach, called Abductive ILP (A/ILP), for machine learning of biological networks from empirical data. It describes the A/ILP approach applied to different biological problems and reviews the main results. These problems are: (1) machine learning of metabolic networks applied to predictive toxicology; (2) multi-clause learning (MCL) of metabolic control points; (3) learning a causal network from temporal gene expression data and (4) automatic construction of probabilistic trophic networks. Hypothetical network structures and parameters generated by A/ILP in each application are assessed in terms of predictive accuracy as well as biological insight provided. The chapter discusses how Progol 5.0 and MC-Toplog have been used for learning biological networks from realworld data. A key advantage of the logical modeling approach here compared with the Bayesian approach is the ability to incorporate relational background knowledge of existing known biochemical pathways, enzyme classes, species information, etc.
Muggleton SH, 2014, Alan Turing and the development of Artificial Intelligence, AI Communications, Vol: 27, Pages: 3-10
Pahlavi N, Muggleton SH, 2014, Towards efficient higher-order logic learning in a first-order datalog framework, Latest Advances in Inductive Logic Programming, Pages: 209-216, ISBN: 9781783265084
Within inductive logic programming (ILP), the concepts to be learned are normally considered as being succinctly representable in first-order logic. In a previous chapter the authors demonstrated that increased predictive accuracy can be achieved by employing higher-order logic (HOL) in the background knowledge. In this chapter, the flexible higher-order Horn clauses (FHOHC) framework is introduced. It is more expressive than the formalism used previously and can be emulated (with the use of “holds” statements and flattening) in a fragment of Datalog. The decidability, compatibility with ILP systems like Progol and positive learnability results of Datalog are then used towards efficient higherorder logic learning (HOLL). We show with experiments that this approach outperforms the HOLL system λProgol and that it can learn concepts in other HOLL settings like learning HOL and using HOL for abduction.
Muggleton SH, Watanabe H, 2014, Latest advances in inductive logic programming, ISBN: 9781783265084
This book represents a selection of papers presented at the Inductive Logic Programming (ILP) workshop held at Cumberland Lodge, Great Windsor Park. The collection marks two decades since the first ILP workshop in 1991. During this period the area has developed into the main forum for work on logic-based machine learning. The chapters cover a wide variety of topics, ranging from theory and ILP implementations to state-of-the-art applications in real-world domains. The international contributors represent leaders in the field from prestigious institutions in Europe, North America and Asia. Graduate students and researchers in this field will find this book highly useful as it provides an up-to-date insight into the key sub-areas of implementation and theory of ILP. For academics and researchers in the field of artificial intelligence and natural sciences, the book demonstrates how ILP is being used in areas as diverse as the learning of game strategies, robotics, natural language understanding, query search, drug design and protein modelling.
Fidjeland AK, Luk W, Muggleton SH, 2014, Customisable multi-processor acceleration of inductive logic programming, Latest Advances in Inductive Logic Programming, Pages: 123-141, ISBN: 9781783265084
Parallel approaches to Inductive Logic Programming (ILP) are adopted to address the computational complexity in the learning process. Existing parallel ILP implementations build on conventional general-purpose processors. This chapter describes a different approach, by exploiting usercustomisable parallelism available in advanced reconfigurable devices such as Field-Programmable Gate Arrays (FPGAs). Our customisable parallel architecture for ILP has three elements: a customisable logic programming processor, a multi-processor for parallel hypothesis evaluation, and an architecture generation framework for creating such multi-processors. Our approach offers a means of achieving high performance by producing parallel architectures adapted both to the problem domain and to specific problem instances. The coverage test in Progol 4.4 is performed up to 56 times faster using our multi-processor.
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.