Results
- Showing results for:
- Reset all filters
Search results
-
Journal articleAlAttar A, Cursi F, Kormushev P, 2021,
Kinematic-model-free redundancy resolution using multi-point tracking and control for robot manipulation
, Applied Sciences, Vol: 11, Pages: 1-15, ISSN: 2076-3417Abstract: Robots have been predominantly controlled using conventional control methods that require prior knowledge of the robots’ kinematic and dynamic models. These controllers can be challenging to tune and cannot directly adapt to changes in kinematic structure or dynamic properties. On the other hand, model-learning controllers can overcome such challenges.Our recently proposed model-learning orientation controller has shown promising ability to simul6 taneously control a three-degrees-of-freedom robot manipulator’s end-effector pose. However, this controller does not perform optimally with robots of higher degrees-of-freedom nor does it resolve redundancies. The research presented in this paper extends the state-of-the-art kinematic9 model-free controller to perform pose control of hyper-redundant robot manipulators and resolve redundancies by tracking and controlling multiple points along the robot’s serial chain. The results show that with more control points, the controller is able to reach desired poses in fewer steps, yielding an improvement of up to 66%, and capable of achieving complex configurations. The algorithm was validated by running the simulation 100 times and it was found that 82% of the times the robot successfully reached the desired target pose within 150 steps.
-
Conference paperDejl A, He P, Mangal P, et al., 2021,
Argflow: a toolkit for deep argumentative explanations for neural networks
, Proc. of the 20th International Conference on Autonomous Agents and Multiagent Systems, Pages: 1761-1763, ISSN: 1558-2914In recent years, machine learning (ML) models have been successfully applied in a variety of real-world applications. However, theyare often complex and incomprehensible to human users. This candecrease trust in their outputs and render their usage in criticalsettings ethically problematic. As a result, several methods for explaining such ML models have been proposed recently, in particularfor black-box models such as deep neural networks (NNs). Nevertheless, these methods predominantly explain outputs in termsof inputs, disregarding the inner workings of the ML model computing those outputs. We present Argflow, a toolkit enabling thegeneration of a variety of ‘deep’ argumentative explanations (DAXs)for outputs of NNs on classification tasks.
-
Journal articleMyall AC, Peach RL, Weiße AY, et al., 2021,
Network memory in the movement of hospital patients carrying drug-resistant bacteria
, Applied Network Science, Vol: 6, ISSN: 2364-8228Hospitals constitute highly interconnected systems that bring into contact anabundance of infectious pathogens and susceptible individuals, thus makinginfection outbreaks both common and challenging. In recent years, there hasbeen a sharp incidence of antimicrobial-resistance amongsthealthcare-associated infections, a situation now considered endemic in manycountries. Here we present network-based analyses of a data set capturing themovement of patients harbouring drug-resistant bacteria across three largeLondon hospitals. We show that there are substantial memory effects in themovement of hospital patients colonised with drug-resistant bacteria. Suchmemory effects break first-order Markovian transitive assumptions andsubstantially alter the conclusions from the analysis, specifically on noderankings and the evolution of diffusive processes. We capture variable lengthmemory effects by constructing a lumped-state memory network, which we then useto identify overlapping communities of wards. We find that these communities ofwards display a quasi-hierarchical structure at different levels of granularitywhich is consistent with different aspects of patient flows related to hospitallocations and medical specialties.
-
Conference paperTavakoli A, Fatemi M, Kormushev P, 2021,
Learning to represent action values as a hypergraph on the action vertices
, Vienna, Austria, International Conference on Learning RepresentationsAction-value estimation is a critical component of many reinforcement learning(RL) methods whereby sample complexity relies heavily on how fast a good estimator for action value can be learned. By viewing this problem through the lens ofrepresentation learning, good representations of both state and action can facilitateaction-value estimation. While advances in deep learning have seamlessly drivenprogress in learning state representations, given the specificity of the notion ofagency to RL, little attention has been paid to learning action representations. Weconjecture that leveraging the combinatorial structure of multi-dimensional actionspaces is a key ingredient for learning good representations of action. To test this,we set forth the action hypergraph networks framework—a class of functions forlearning action representations in multi-dimensional discrete action spaces with astructural inductive bias. Using this framework we realise an agent class basedon a combination with deep Q-networks, which we dub hypergraph Q-networks.We show the effectiveness of our approach on a myriad of domains: illustrativeprediction problems under minimal confounding effects, Atari 2600 games, anddiscretised physical control benchmarks.
-
Journal articleEspinosa-Gonzalez AB, Neves AL, Fiorentino F, et al., 2021,
Predicting Risk of Hospital Admission in Patients With Suspected COVID-19 in a Community Setting: Protocol for Development and Validation of a Multivariate Risk Prediction Tool
, JMIR RESEARCH PROTOCOLS, Vol: 10, ISSN: 1929-0748- Author Web Link
- Cite
- Citations: 8
-
Journal articlePeach RL, Arnaudon A, Schmidt JA, et al., 2021,
HCGA: Highly comparative graph analysis for network phenotyping
, Patterns, Vol: 2, Pages: 100227-100227, ISSN: 2666-3899<jats:title>A<jats:sc>bstract</jats:sc></jats:title><jats:p>Networks are widely used as mathematical models of complex systems across many scientific disciplines, not only in biology and medicine but also in the social sciences, physics, computing and engineering. Decades of work have produced a vast corpus of research characterising the topological, combinatorial, statistical and spectral properties of graphs. Each graph property can be thought of as a feature that captures important (and some times overlapping) characteristics of a network. In the analysis of real-world graphs, it is crucial to integrate systematically a large number of diverse graph features in order to characterise and classify networks, as well as to aid network-based scientific discovery. In this paper, we introduce HCGA, a framework for highly comparative analysis of graph data sets that computes several thousands of graph features from any given network. HCGA also offers a suite of statistical learning and data analysis tools for automated identification and selection of important and interpretable features underpinning the characterisation of graph data sets. We show that HCGA outperforms other methodologies on supervised classification tasks on benchmark data sets whilst retaining the interpretability of network features. We also illustrate how HCGA can be used for network-based discovery through two examples where data is naturally represented as graphs: the clustering of a data set of images of neuronal morphologies, and a regression problem to predict charge transfer in organic semiconductors based on their structure. HCGA is an open platform that can be expanded to include further graph properties and statistical learning tools to allow researchers to leverage the wide breadth of graph-theoretical research to quantitatively analyse and draw insights from network data.</jats:p>
-
Journal articleTajnafoi G, Arcucci R, Mottet L, et al., 2021,
Variational Gaussian process for optimal sensor placement
, Applications of Mathematics, Vol: 66, Pages: 287-317, ISSN: 0373-6725Sensor placement is an optimisation problem that has recently gained great relevance. In order to achieve accurate online updates of a predictive model, sensors are used to provide observations. When sensor location is optimally selected, the predictive model can greatly reduce its internal errors. A greedy-selection algorithm is used for locating these optimal spatial locations from a numerical embedded space. A novel architecture for solving this big data problem is proposed, relying on a variational Gaussian process. The generalisation of the model is further improved via the preconditioning of its inputs: Masked Autoregressive Flows are implemented to learn nonlinear, invertible transformations of the conditionally modelled spatial features. Finally, a global optimisation strategy extending the Mutual Information-based optimisation and fine-tuning of the selected optimal location is proposed. The methodology is parallelised to speed up the computational time, making these tools very fast despite the high complexity associated with both spatial modelling and placement tasks. The model is applied to a real three-dimensional test case considering a room within the Clarence Centre building located in Elephant and Castle, London, UK.
-
Journal articleSivan M, Rayner C, Delaney B, 2021,
Fresh evidence of the scale and scope of long covid
, BMJ-BRITISH MEDICAL JOURNAL, Vol: 373, ISSN: 0959-535X- Author Web Link
- Cite
- Citations: 34
-
Journal articleRussell F, Takeda Y, Kormushev P, et al., 2021,
Stiffness modulation in a humanoid robotic leg and knee
, IEEE Robotics and Automation Letters, Vol: 6, Pages: 2563-2570, ISSN: 2377-3766Stiffness modulation in walking is critical to maintain static/dynamic stability as well as minimize energy consumption and impact damage. However, optimal, or even functional, stiffness parameterization remains unresolved in legged robotics.We introduce an architecture for stiffness control utilizing a bioinspired robotic limb consisting of a condylar knee joint and leg with antagonistic actuation. The joint replicates elastic ligaments of the human knee providing tuneable compliance for walking. It further locks out at maximum extension, providing stability when standing. Compliance and friction losses between joint surfaces are derived as a function of ligament stiffness and length. Experimental studies validate utility through quantification of: 1) hip perturbation response; 2) payload capacity; and 3) static stiffness of the leg mechanism.Results prove initiation and compliance at lock out can be modulated independently of friction loss by changing ligament elasticity. Furthermore, increasing co-contraction or decreasing joint angle enables increased leg stiffness, which establishes co-contraction is counterbalanced by decreased payload.Findings have direct application in legged robots and transfemoral prosthetic knees, where biorobotic design could reduce energy expense while improving efficiency and stability. Future targeted impact involves increasing power/weight ratios in walking robots and artificial limbs for increased efficiency and precision in walking control.
-
Journal articleWu P, Chang X, Yuan W, et al., 2021,
Fast data assimilation (FDA): Data assimilation by machine learning for faster optimize model state
, JOURNAL OF COMPUTATIONAL SCIENCE, Vol: 51, ISSN: 1877-7503- Author Web Link
- Cite
- Citations: 6
-
Journal articleCyras K, Heinrich Q, Toni F, 2021,
Computational complexity of flat and generic assumption-based argumentation, with and without probabilities
, Artificial Intelligence, Vol: 293, Pages: 1-36, ISSN: 0004-3702Reasoning with probabilistic information has recently attracted considerable attention in argumentation, and formalisms of Probabilistic Abstract Argumentation (PAA), Probabilistic Bipolar Argumentation (PBA) and Probabilistic Structured Argumentation (PSA) have been proposed. These foundational advances have been complemented with investigations on the complexity of some approaches to PAA and PBA, but not to PSA. We study the complexity of an existing form of PSA, namely Probabilistic Assumption-Based Argumentation (PABA), a powerful, implemented formalism which subsumes several forms of PAA and other forms of PSA. Specifically, we establish membership (general upper bounds) and completeness (instantiated lower bounds) of reasoning in PABA for the class FP#P (of functions with a #P-oracle for counting the solutions of an NP problem) with respect to newly introduced probabilistic verification, credulous and sceptical acceptance function problems under several ABA semantics. As a by-product necessary to establish PABA complexity results, we provide a comprehensive picture of the ABA complexity landscape (for both flat and generic, possibly non-flat ABA) for the classical decision problems of verification, existence, credulous and sceptical acceptance under those ABA semantics.
-
Journal articleQian Y, Expert P, Panzarasa P, et al., 2021,
Geometric graphs from data to aid classification tasks with Graph Convolutional Networks
, Patterns, Vol: 2, Pages: 100237-100237, ISSN: 2666-3899 -
Conference paperBonavita M, Arcucci R, Carrassi A, et al., 2021,
Machine Learning for Earth System Observation and Prediction
, Publisher: AMER METEOROLOGICAL SOC, Pages: E710-E716, ISSN: 0003-0007- Author Web Link
- Cite
- Citations: 17
-
Journal articleMaes A, Barahona M, Clopath C, 2021,
Learning compositional sequences with multiple time scales through a hierarchical network of spiking neurons
, PLoS Computational Biology, Vol: 17, ISSN: 1553-734XSequential behaviour is often compositional and organised across multiple time scales: a set of individual elements developing on short time scales (motifs) are combined to form longer functional sequences (syntax). Such organisation leads to a natural hierarchy that can be used advantageously for learning, since the motifs and the syntax can be acquired independently. Despite mounting experimental evidence for hierarchical structures in neuroscience, models for temporal learning based on neuronal networks have mostly focused on serial methods. Here, we introduce a network model of spiking neurons with a hierarchical organisation aimed at sequence learning on multiple time scales. Using biophysically motivated neuron dynamics and local plasticity rules, the model can learn motifs and syntax independently. Furthermore, the model can relearn sequences efficiently and store multiple sequences. Compared to serial learning, the hierarchical model displays faster learning, more flexible relearning, increased capacity, and higher robustness to perturbations. The hierarchical model redistributes the variability: it achieves high motif fidelity at the cost of higher variability in the between-motif timings.
-
Journal articleEspinosa-Gonzalez AB, Neves AL, Fiorentino F, et al., 2021,
Predicting Risk of Hospital Admission in Patients With Suspected COVID-19 in a Community Setting: Protocol for Development and Validation of a Multivariate Risk Prediction Tool (Preprint)
<sec> <title>BACKGROUND</title> <p>During the pandemic, remote consultations have become the norm for assessing patients with signs and symptoms of COVID-19 to decrease the risk of transmission. This has intensified the clinical uncertainty already experienced by primary care clinicians when assessing patients with suspected COVID-19 and has prompted the use of risk prediction scores, such as the National Early Warning Score (NEWS2), to assess severity and guide treatment. However, the risk prediction tools available have not been validated in a community setting and are not designed to capture the idiosyncrasies of COVID-19 infection.</p> </sec> <sec> <title>OBJECTIVE</title> <p>The objective of this study is to produce a multivariate risk prediction tool, RECAP-V1 (Remote COVID-19 Assessment in Primary Care), to support primary care clinicians in the identification of those patients with COVID-19 that are at higher risk of deterioration and facilitate the early escalation of their treatment with the aim of improving patient outcomes.</p> </sec> <sec> <title>METHODS</title> <p>The study follows a prospective cohort observational design, whereby patients presenting in primary care with signs and symptoms suggestive of COVID-19 will be followed and their data linked to hospital outcomes (hospital admission and death). Data collection will be carried out by primary care clinicians in four arms: North West London Clinical Commissioning Groups (NWL CCGs), Oxford-Royal College of General Practitioners (RCGP) Research and Surveillance Centre (RSC), Covid Clinical Assessment Service (CCAS), and South East London CCGs (Doctaly platform). The study involves the use o
-
Journal articleCursi F, Modugno V, Lanari L, et al., 2021,
Bayesian neural network modeling and hierarchical MPC for a tendon-driven surgical robot with uncertainty minimization
, IEEE Robotics and Automation Letters, Vol: 6, Pages: 2642-2649, ISSN: 2377-3766In order to guarantee precision and safety in robotic surgery, accurate models of the robot and proper control strategies are needed. Bayesian Neural Networks (BNN) are capable of learning complex models and provide information about the uncertainties of the learned system. Model Predictive Control (MPC) is a reliable control strategy to ensure optimality and satisfaction of safety constraints. In this work we propose the use of BNN to build the highly nonlinear kinematic and dynamic models of a tendon-driven surgical robot, and exploit the information about the epistemic uncertainties by means of a Hierarchical MPC (Hi-MPC) control strategy. Simulation and real world experiments show that the method is capable of ensuring accurate tip positioning, while satisfying imposed safety bounds on the kinematics and dynamics of the robot.
-
Journal articleCheng S, Pain CC, Guo Y-K, et al., 2021,
Real-time Updating of Dynamic Social Networks for COVID-19 Vaccination Strategies
<jats:title>Abstract</jats:title><jats:p>Vaccination strategy is crucial in fighting the COVID-19 pandemic. Since the supply is still limited in many countries, contact network-based interventions can be most powerful to set an efficient strategy by identifying high-risk individuals or communities. However, due to the high dimension, only partial and noisy network information can be available in practice, especially for dynamic systems where contact networks are highly time-variant. Furthermore, the numerous mutations of SARS-CoV-2 have a significant impact on the infectious probability, requiring real-time network updating algorithms. In this study, we propose a sequential network updating approach based on data assimilation techniques to combine different sources of temporal information. We then prioritise the individuals with high-degree or high-centrality, obtained from assimilated networks, for vaccination. The assimilation-based approach is compared with the standard method (based on partially observed networks) and a random selection strategy in terms of vaccination effectiveness in a SIR model. The numerical comparison is first carried out using real-world face-to-face dynamic networks collected in a high school, followed by sequential multi-layer networks generated relying on the Barabasi-Albert model emulating large-scale social networks with several communities.</jats:p>
-
Journal articleKostopoulou O, Tracey C, Delaney B, 2021,
Can decision support combat incompleteness and bias in routine primary care data?
, Journal of the American Medical Informatics Association, ISSN: 1067-5027Objective: Routine primary care data may be used for the derivation of clinical prediction rules and risk scores. We sought to measure the impact of a decision support system (DSS) on data completeness and freedom from bias.Materials and Methods: We used the clinical documentation of 34 UK general practitioners who took part in a previous study evaluating the DSS. They consulted with 12 standardized patients. In addition to suggesting di- agnoses, the DSS facilitates data coding. We compared the documentation from consultations with the elec- tronic health record (EHR) (baseline consultations) vs consultations with the EHR-integrated DSS (supported consultations). We measured the proportion of EHR data items related to the physician’s final diagnosis. We expected that in baseline consultations, physicians would document only or predominantly observations re- lated to their diagnosis, while in supported consultations, they would also document other observations as a re- sult of exploring more diagnoses and/or ease of coding.Results: Supported documentation contained significantly more codes (incidence rate ratio [IRR] 1⁄4 5.76 [4.31, 7.70] P < .001) and less free text (IRR 1⁄4 0.32 [0.27, 0.40] P < .001) than baseline documentation. As expected, the proportion of diagnosis-related data was significantly lower (b 1⁄4 0.08 [0.11, 0.05] P < .001) in the supported consultations, and this was the case for both codes and free text.Conclusions: We provide evidence that data entry in the EHR is incomplete and reflects physicians’ cognitive biases. This has serious implications for epidemiological research that uses routine data. A DSS that facilitates and motivates data entry during the consultation can improve routine documentation.
-
Journal articleZheng JX, Pawar S, Goodman DFM, 2021,
Further towards unambiguous edge bundling: Investigating power-confluentdrawings for network visualization
, IEEE Transactions on Visualization and Computer Graphics, Vol: 27, Pages: 2244-2249, ISSN: 1077-2626Bach et al. [1] recently presented an algorithm for constructing confluentdrawings, by leveraging power graph decomposition to generate an auxiliaryrouting graph. We identify two problems with their method and offer a singlesolution to solve both. We also classify the exact type of confluent drawingsthat the algorithm can produce as 'power-confluent', and prove that it is asubclass of the previously studied 'strict confluent' drawing. A descriptionand source code of our implementation is also provided, which additionallyincludes an improved method for power graph construction.
-
Journal articleSaputra RP, Rakicevic N, Chappell D, et al., 2021,
Hierarchical decomposed-objective model predictive control for autonomous casualty extraction
, IEEE Access, Vol: 9, Pages: 39656-39679, ISSN: 2169-3536In recent years, several robots have been developed and deployed to perform casualty extraction tasks. However, the majority of these robots are overly complex, and require teleoperation via either a skilled operator or a specialised device, and often the operator must be present at the scene to navigate safely around the casualty. Instead, improving the autonomy of such robots can reduce the reliance on expert operators and potentially unstable communication systems, while still extracting the casualty in a safe manner. There are several stages in the casualty extraction procedure, from navigating to the location of the emergency, safely approaching and loading the casualty, to finally navigating back to the medical assistance location. In this paper, we propose a Hierarchical Decomposed-Objective based Model Predictive Control (HiDO-MPC) method for safely approaching and manoeuvring around the casualty. We implement this controller on ResQbot — a proof-of-concept mobile rescue robot we previously developed — capable of safely rescuing an injured person lying on the ground, i.e. performing the casualty extraction procedure. HiDO-MPC achieves the desired casualty extraction behaviour by decomposing the main objective into multiple sub-objectives with a hierarchical structure. At every time step, the controller evaluates this hierarchical decomposed objective and generates the optimal control decision. We have conducted a number of experiments both in simulation and using the real robot to evaluate the proposed method’s performance, and compare it with baseline approaches. The results demonstrate that the proposed control strategy gives significantly better results than baseline approaches in terms of accuracy, robustness, and execution time, when applied to casualty extraction scenarios.
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.
Contact us
Artificial Intelligence Network
South Kensington Campus
Imperial College London
SW7 2AZ
To reach the elected speaker of the network, Dr Rossella Arcucci, please contact:
To reach the network manager, Diana O'Malley - including to join the network - please contact: