More

Dr Sarah Greenfield

Job: Research Fellow

Faculty: Computing, Engineering and Media

School/department: School of Computer Science and Informatics

Research group(s): The De Montfort University Interdisciplinary Group in Intelligent Transport Systems (DIGITS), Centre for Computational Intelligence (CCI)

Address: De Montfort University, The Gateway, Leicester, LE1 9BH

T: +44 (0)116 250 6171

E: s.greenfield@dmu.ac.uk

W: https://www.dmu.ac.uk/digits

 

Close all sections
Open all sections

Personal profile

Sarah Greenfield received the BA in Mathematics and Philosophy from London University in 1978. In 2005 she was awarded a distinction in the MSc IT degree from De Montfort University, Leicester, UK, and in 2012 she was awarded a PhD at De Montfort's Centre for Computational Intelligence, working under the supervision of Prof Chiclana. Her enduring interest in logic and the philosophy of mathematics was reflected in her original choice of degree subject. Her MSc project was in the field of type-2 fuzzy logic, and her PhD studies continued this theme in her exploration of mathematical and philosophical aspects of type-2 fuzzy logic in relation to such topics as uncertainty modelling and defuzzification.   Since completing her studies she has widened her research interests to include complex fuzzy inferencing and computational intelligence in transport.

Research group affiliations

The De Montfort University Interdisciplinary Group in Intelligent Transport Systems (DIGITS)

Centre for Computational Intelligence (CCI)  

Publications and outputs

  • A Fuzzy Prescreening Tool to Assist in the Diagnosis of High Functioning Individuals on the Autism Spectrum Who Present with Mental Health Comorbidities
    dc.title: A Fuzzy Prescreening Tool to Assist in the Diagnosis of High Functioning Individuals on the Autism Spectrum Who Present with Mental Health Comorbidities dc.contributor.author: Smith, Philip; Greenfield, Sarah dc.description.abstract: Autism Spectrum Disorder is a neurological developmental disorder that effects at least 1% of the population, the majority of cases are high functioning individuals who struggle to get positive diagnoses that are vital to obtain community support. In this study, we have created and tested a Fuzzy Inferencing System to support clinicians, psychologists, family members and relevant stake holders to increase the chances for high functioning individuals to get a referral for full assessment to determine an autism diagnosis.
  • Out-of-Distribution Detection with Memory-Augmented Variational Autoencoder
    dc.title: Out-of-Distribution Detection with Memory-Augmented Variational Autoencoder dc.contributor.author: Ataeiasad, Faezeh; Elizondo, David; Ramírez, Saúl Calderón; Greenfield, Sarah; Deka, Lipika dc.description.abstract: This paper proposes a novel method capable of both detecting OOD data and generating in-distribution data samples. To achieve this, a VAE model is adopted and augmented with a memory module, providing capacities for identifying OOD data and synthesising new in-distribution samples. The proposed VAE is trained on normal data and the memory stores prototypical patterns of the normal data distribution. At test time, the input is encoded by the VAE encoder; this encoding is used as a query to retrieve related memory items, which are then integrated with the input encoding and passed to the decoder for reconstruction. Normal samples reconstruct well and yield low reconstruction errors, while OOD inputs produce high reconstruction errors as their encodings get replaced by retrieved normal patterns. Prior works use memory modules for OOD detection with autoencoders, but this method leverages a VAE architecture to enable generation abilities. Experiments conducted with CIFAR-10 and MNIST datasets show that the memory-augmented VAE consistently outperforms the baseline, particularly where OOD data resembles normal patterns. This notable improvement is due to the enhanced latent space representation provided by the VAE. Overall, the memory-equipped VAE framework excels in identifying OOD and generating creative examples effectively. dc.description: open access article
  • Towards Refined Autism Screening: A Fuzzy Logic Approach with a Focus on Subtle Diagnostic Challenges
    dc.title: Towards Refined Autism Screening: A Fuzzy Logic Approach with a Focus on Subtle Diagnostic Challenges dc.contributor.author: Smith, Philip; Greenfield, Sarah dc.description.abstract: This study explores the creation and testing of a Fuzzy Inferencing System for automating preliminary referrals for autism diagnosis, utilizing membership functions aligned with the Autism Quotient 10-item questionnaire. Validated across three distinct datasets, the system demonstrated perfect accuracy in deterministic settings and an overall accuracy of 92.91% in a broad fuzzy dataset. The use of Fuzzy Logic reflects the complex and variable nature of autism diagnosis, suggesting its potential applicability in this field. While the system effectively categorized clear referral and non-referral scenarios, it faced challenges in accurately identifying cases requiring a second opinion. These results indicate the need for further refinement to enhance the efficiency and accuracy of preliminary autism screenings, pointing to future avenues for improving the system’s performance. The motivation behind this study is to address the diagnostic gap for high-functioning adults whose symptoms present in a more neurotypical manner. Many current deep learning approaches for diagnosing autism focus on quantitative datasets like fMRI and facial expressions, often overlooking behavioral traits. However, autism diagnosis still heavily relies on long histories and multi-stakeholder information from parents, teachers, doctors and behavioral experts. This research addresses the challenge of creating an automated system that can handle the nuances and variability inherent in ASD symptoms. The theoretical innovation lies in the novel application of Fuzzy Logic to interpret these subtle diagnostic indicators, providing a more systematic approach compared to traditional methods. By bridging the gap between subjective clinical evaluations and objective computational techniques, this study aims to enhance the preliminary screening process for ASD. dc.description: open access article
  • The Stratic Defuzzifier for Discretised General Type-2 Fuzzy Sets
    dc.title: The Stratic Defuzzifier for Discretised General Type-2 Fuzzy Sets dc.contributor.author: Greenfield, Sarah; Chiclana, Francisco dc.description.abstract: Stratification is a feature of the type-reduced set of the general type-2 fuzzy set, from which a new technique for general type-2 defuzzification, Stratic Defuzzification, may be derived. Existing defuzzification strategies are summarised. The stratified structure is described, after which the Stratic Defuzzifier is presented and contrasted experimentally for accuracy and efficiency with both the Exhaustive Method of Defuzzification (to benchmark accuracy) and the alpha-Planes/Karnik–Mendel Iterative Procedure strategy, employing 5, 11, 21, 51 and 101 alpha-planes. The Stratic Defuzzifier is shown to be much faster than the Exhaustive Defuzzifier. In fact the Stratic Defuzzifier and the alpha-Planes/Karnik–Mendel Iterative Procedure Method are comparably speedy; the speed of execution correlates with the number of planes participating in the defuzzification process. The accuracy of the Stratic Defuzzifier is shown to be excellent. It is demonstrated to be more accurate than the alpha-Planes/Karnik–Mendel Iterative Procedure Method in four of six test cases, regardless of the number of -planes employed. In one test case, it is less accurate than the alpha-Planes/Karnik–Mendel Iterative Procedure Method, regardless of the number of alpha-planes employed. In the remaining test case, the alpha-Planes/Karnik–Mendel Iterative Procedure Method with 11 alpha-Planes gives the most accurate result, with the Stratic Defuzzifier coming second. dc.description: The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.
  • The Collapsing Defuzzifier for discretised generalised type-2 fuzzy sets
    dc.title: The Collapsing Defuzzifier for discretised generalised type-2 fuzzy sets dc.contributor.author: Greenfield, Sarah; Chiclana, Francisco dc.description.abstract: The Greenfield–Chiclana Collapsing Defuzzifier is an established efficient accurate technique for the defuzzification of the interval type-2 fuzzy set. This paper reports on the extension of the Collapsing Defuzzifier to the generalised type-2 fuzzy set. Existing techniques for the defuzzification of generalised type-2 fuzzy sets are presented after which the interval Collapsing Defuzzifier is summarised. The collapsing technique is then extended to generalised type-2 fuzzy sets, giving the Generalised Greenfield–Chiclana Collapsing Defuzzifier. This is contrasted experimentally with both the benchmark Exhaustive Defuzzifier and the α-Planes/Karnik–Mendel Iterative Procedure approach in relation to efficiency and accuracy. The GGCCD is demonstrated to be many times faster than the Exhaustive Defuzzifier and its accuracy is shown to be excellent. In relation to the α-Planes/Karnik–Mendel Iterative Procedure approach it is shown to be comparable in accuracy, but faster. dc.description: The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.
  • Geometric Defuzzification Revisited
    dc.title: Geometric Defuzzification Revisited dc.contributor.author: Greenfield, Sarah dc.description.abstract: In this paper the Geometric Defuzzification strategy for type-2 fuzzy sets is reappraised. For both discretised and geometric fuzzy sets the techniques for type-1, interval type-2, and generalised type-2 defuzzification are presented in turn. In the type-2 case the accuracy of Geometric Defuzzification is assessed through a series of test runs on interval type-2 fuzzy sets, using Exhaustive Defuzzification as the benchmark method. These experiments demonstrate the Geometric Defuzzifier to be wildly inaccurate. The test sets take many shapes; they are not confined to those type-2 sets with rotational symmetry that have previously been acknowledged by the technique’s developers to be problematic as regards accuracy. Type-2 Geometric Defuzzification is then examined theoretically. The defuzzification strategy is demonstrated to be built upon a fallacious application of the concept of centroid. This explains the markedly inaccurate experimental results. Thus the accuracy issues of type-2 Geometric Defuzzification are revealed to be inevitable, fundamental and significant. dc.description: The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.
  • Type-Reduced Set Structure and the Truncated Type-2 Fuzzy Set
    dc.title: Type-Reduced Set Structure and the Truncated Type-2 Fuzzy Set dc.contributor.author: Greenfield, Sarah; Chiclana, Francisco dc.description.abstract: In this paper, the Type-Reduced Set (TRS) of the continuous type-2 fuzzy set is considered as an object in its own right. The structures of the TRSs of both the interval and generalised forms of the type-2 fuzzy set are investigated. In each case the respective TRS structure is approached by first examining the TRS of the discretised set. The TRS of a continuous interval type-2 fuzzy set is demonstrated to be a continuous horizontal straight line, and that of a generalised type-2 fuzzy set, a continuous, convex curve. This analysis leads on to the concept of truncation, and the definition of the truncation grade. The truncated type-2 fuzzy set is then defined, whose TRS (and hence defuzzified value) is identical to that of the non-truncated type-2 fuzzy set. This result is termed the Type-2 Truncation Theorem, an immediate corollary of which is the Type-2 Equivalence Theorem which states that the defuzzified values of type-2 fuzzy sets that are equivalent under truncation are equal. Experimental corroboration of the equivalence of the non-truncated and truncated generalised type-2 fuzzy set is provided. The implications of these theorems for uncertainty quantification are explored. The theorem’s repercussions for type-2 defuzzification employing the α-Planes Representation are examined; it is shown that the known inaccuracies of the α-Planes Method are deeply entrenched. dc.description: The file attached to this record is the author's final peer reviewed version.
  • Join and Meet Operations for Interval-Valued Complex Fuzzy Logic
    dc.title: Join and Meet Operations for Interval-Valued Complex Fuzzy Logic dc.contributor.author: Greenfield, Sarah; Chiclana, Francisco; Dick, Scott dc.description.abstract: Interval-valued complex fuzzy logic is able to handle scenarios where both seasonality and uncertainty feature. The interval-valued complex fuzzy set is defined, and the interval valued complex fuzzy inferencing system outlined. Highly pertinent to complex fuzzy logic operations is the concept of rotational invariance, which is an intuitive and desirable characteristic. Interval-valued complex fuzzy logic is driven by interval-valued join and meet operations. Four pairs of alternative algorithms for these operations are specified; three pairs possesses the attribute of rotational invariance, whereas the other pair lacks this characteristic. dc.description: DMU Interdisciplinary Group in Intelligent Transport Systems
  • Uncertainty Measurement for the Interval Type-2 Fuzzy Set
    dc.title: Uncertainty Measurement for the Interval Type-2 Fuzzy Set dc.contributor.author: Greenfield, Sarah dc.description.abstract: In this paper, two measures of uncertainty for interval type-2 fuzzy sets are presented, evaluated, compared and contrasted. Wu and Mendel regard the length of the type-reduced set as a measure of the uncertainty in an interval set. Green eld and John argue that the volume under the surface of the type-2 fuzzy set is a measure of the uncertainty relating to the set. For an interval type-2 fuzzy set, the volume measure is equivalent to the area of the footprint of uncertainty of the set. Experiments show that though the two measures give di erent results, there is considerable commonality between them. The concept of invariance under vertical translation is introduced; the uncertainty measure of a fuzzy set has the property of invariance under vertical translation if the value it generates remains constant under any vertical translation of the fuzzy set. It is left unresolved whether invariance under vertical translation is an essential property of a type-2 uncertainty measure.
  • Interval-Valued Complex Fuzzy Logic
    dc.title: Interval-Valued Complex Fuzzy Logic dc.contributor.author: Greenfield, Sarah; Chiclana, Francisco; Dick, Scott dc.description.abstract: Data is frequently characterised by both uncertainty and seasonality. Type-2 fuzzy sets are an extension of type-1 fuzzy sets offering a conceptual scheme within which the effects of uncertainties in fuzzy inferencing may be modelled and minimised. Complex fuzzy sets are type-1 fuzzy sets extended by an additional phase term which permits them to intuitively represent the seasonal aspect of fuzziness in time-series applications. Type-2 fuzzy sets take two forms, generalised, and the simpler interval. Interval-valued fuzzy sets are type-1 fuzzy sets whose behaviour and properties are equivalent to interval type-2 fuzzy sets. This paper is concerned with the combination of interval-valued fuzzy sets and complex fuzzy sets to develop interval-valued complex fuzzy sets, an adaption of complex fuzzy sets such that the membership function assigns each point on the domain to an interval. From the definition of the interval-valued complex fuzzy set, the principles of interval-valued complex fuzzy logic are developed. dc.description: DMU Interdisciplinary Group in Intelligent Transport Systems (DIGITS)

Click here to view a full listing of Sarah Greenfield's publications and outputs.

Key research outputs

[1] Sarah Greenfield, Francisco Chiclana, Simon Coupland and Robert I. John, “The collapsing method of defuzzification for discretised interval type-2 fuzzy sets”, Information Sciences Special Issue, “High Order Fuzzy Sets: Theory and Applications”, volume 179, issue 13, pages 2055–2069, June 2009. DOI: http://dx.doi.org/10.1016/j.ins.2008.07.011; ISSN: 0020–0255.

[2] Sarah Greenfield and Francisco Chiclana, “Type-Reduction of the Discretised Interval Type-2 Fuzzy Set: Approaching the Continuous Case through Progressively Finer Discretisation”, Journal of Artificial Intelligence and Soft Computing Research, volume 1, issue 3, pages 183–193, 2011. ISSN: 2083–2567.

[3] Sarah Greenfield, Francisco Chiclana, Robert I. John and Simon Coupland, “The sampling method of defuzzification for type-2 fuzzy sets: Experimental evaluation”, Information Sciences, volume 189, pages 77–92, April 2012. DOI: http://dx.doi.org/10.1016/j.ins.2011.11.042; ISSN: 0020–0255.

[4] Sarah Greenfield and Francisco Chiclana, “Accuracy and complexity evaluation of defuzzification strategies for the discretised interval type-2 fuzzy set”, The International Journal of Approximate Reasoning, volume 54, issue 8, pages 1013–1033, October 2013. DOI: http://dx.doi.org/10.1016/j.ijar.2013.04.013; ISSN: 0888–613X.

[5] Sarah Greenfield and Francisco Chiclana, “Defuzzification of the discretised generalised type-2 fuzzy set: Experimental evaluation”, Information Sciences, volume 244, pages 1–25, September 2013. DOI: http://dx.doi.org/10.1016/j.ins.2013.04.032; ISSN: 0020–0255.

Research interests/expertise

  • Logic
  • Fuzzy logic, including type-2 fuzzy logic and complex fuzzy logic
  • Defuzzification of type-2 fuzzy sets
  • Uncertainty modelling using type-2 fuzzy sets
  • Computational Intelligence in transport

Areas of teaching

  • IMAT 1205: Mathematics for Scientific Computing (Year 1 BSc CGP and AIR)
  • IMAT 3451: Final Year Computing Project supervision

Qualifications

  • PhD in Computational Intelligence, De Montfort University, Leicester, 2012;
  • MSc in Information Technology (Distinction), De Montfort University, Leicester, 2005;
  • HNC (BTEC) Software Engineering Design, Brighton College of Technology, 1989;
  • BA Mathematics & Philosophy (2 II Hons.), Kings College, University of London 1978;
  • A-levels in Mathematics (B), Physics (A) and Chemistry (B), Withington Girls’ School, Manchester 1975.

Courses taught

  • IMAT 1205: Mathematics for Scientific Computing (Year 1 BSc CGP and AIR)
  • IMAT 3451: Final Year Computing Project supervision
  • IMAT 2800: Artificial Intelligence and Modelling for Games (2011 - 2013)

Honours and awards

Anita Borg Scholarship In 2010 I reached the final of the Google Anita Borg Memorial Scholarship: Europe, the Middle East and Africa (www.google.com/anitaborg/emea). The scholarship aims to encourage women to excel in computing and technology, and become active role models and leaders. Awards are based on the strength of candidates’ academic performance, leadership experience and demonstrated passion for computer science. My prize as a finalist was to attend a networking retreat at Google’s Engineering Centre in Zurich.

Creative Thinking Award In 2010 I received the third prize of £2,000 in De Montfort University’s Creative Thinking Awards for my Collapsing Defuzzifier. In my submission for this award I argued that mathematics is a creative enterprise.

Membership of professional associations and societies

  • Member of the European Society for Fuzzy Logic and Technology (EUSFLAT)
  • Member of the Institute of Electrical and Electronics Engineers (IEEE)

Conference attendance

UKCI, September 2005, London, “A Novel Sampling Method for Type-2 Defuzzification”, Sarah Greenfield, Robert I. John and Simon Coupland, presentation. 

FUZZ-IEEE, July 2007, London, “Optimised Generalised Type-2 Join and Meet Operations”, Sarah Greenfield and Robert I. John, presentation.

UKCI, July 2007, London, “The Collapsing Method of Defuzzification for Discretised Interval Type-2 Fuzzy Sets”, Sarah Greenfield, Francisco Chiclana, Robert I. John and Simon Coupland, presentation. 

IPMU 2008, June 2008, Malaga, “Stratification in the Type-Reduced Set and the Generalised Karnik-Mendel Iterative Procedure”, Sarah Greenfield and Robert I. John, presentation.

IFSA-EUSFLAT, July 2009, Lisbon, “The Collapsing Method: Does the Direction of Collapse Affect Accuracy?”, Sarah Greenfield, Francisco Chiclana and Robert I. John, presentation. 

IEEE Symposium on Advances in Type-2 Fuzzy Logic Systems, April 2011, Paris, “Type-Reduction of the Discretised Interval Type-2 Fuzzy Set: What Happens as Discretisation Becomes Finer?”, Sarah Greenfield and Francisco Chiclana, presentation.

EUSFLAT-LFA, July 2011, Aix-Les-Bains, France, “Combining the alpha-Plane Representation with an Interval Defuzzification Method”, Sarah Greenfield and Francisco Chiclana, presentation. 

UKCI 2012, September 2012, Edinburgh, “The Grid Method of Discretisation for Type-2 Fuzzy Sets”, Sarah Greenfield, poster.

EUSFLAT 2013, September 2013, Milan, Italy, “The Structure of the Type-Reduced Set of a Continuous Type-2 Fuzzy Set”, Sarah Greenfield and Francisco Chiclana, presentation.

Externally funded research grants information

Evolutionary Computation for Optimised Rail Travel (EsCORT),  funded by Transport iNet, a research and development project running from 11/11/13 to 31/12/14.  Collaborators are Go Travel Solutions, the Rail Safety and Standards Board, and Network Rail.

Professional esteem indicators

Reviewer for:

  • Applied Soft Computing,
  • IEEE Transactions on Fuzzy Systems,
  • Soft Computing,
  • Journal of Intelligent and Fuzzy Systems,
  • Fuzzy Sets and Systems,
  • Information Sciences, and
  • The International Journal of Approximate Reasoning.

On the programme committee for FCTA 2013 and FCTA 2014. (Fuzzy Computation Theory and Applications).

Co-organiser of a special session on type-2 fuzzy logic at IFSA-EUSFLAT 2009, Lisbon.