Aggregation Operators special session at FUZZIEEE 2014

We are inviting submissions to our proposed special session on Aggregation Operators at FUZZIEEE to be held in Beijing from July 6-14, 2014 as part of the World Congress on Computational Intelligence.

Aggregation Operators (AGOPs) play a key role in fuzzy sets theory as fuzzy logic connectives, and also have wide-ranging applications in decision making, classification and data analysis. Traditional aggregation operators such as the arithmetic mean and median are now acknowledged as particular cases of more general families of aggregation operations, such as the ordered weighted averaging (OWA) operator and Choquet integral. Triangular norms and conorms, uninorms, symmetric sums, to name a few, are widely used families of AGOP. With the growing need to deal with large amounts of data and uncertainty, research on practical applications and their associated challenges is of increasing interest.

This special session will be dedicated to theoretical and practical aspects of AGOPs. Specific topics include, but are not limited to:
- practical constructions of AGOPs,
- real-world applications,
- parameter and weights identification,
- weighting functions for AGOPs,
- AGOPs with specific properties,
- interval-valued and intuitionistic AGOPs,
- theoretical analysis,
- fuzzy measures and integrals.

Style and paper submission guidelines can be found here.

Papers can be submitted here.

Please select “S34: FZ34: Aggregation Operators” from the drop-down box where it says Main research topic.

The current due date for submission is December 20, 2013 – although there is usually a two week extension on this.

Survey on human judgements of species diversity

deakin_logo

We are currently conducting a research project that looks at human perceptions of ecological diversity and how such perceptions feed into judgements regarding how people rank communities for conservation.

The survey is core to our investigation and involves participants making intuitive judgements about the relative diversity of different communities and using the available information to rank the same communities in terms of their conservation importance. The survey takes approximately 15 minutes to complete and it is not necessary for you to be an expert in this area to participate; we hope to have respondents from a range of backgrounds.

More information can be found in the plain language that precedes the survey. The following link will take you to the survey.

https://www.surveymonkey.com/s/biodiversity-project-2

AudioSlides for Knowledge-based systems article

Elsevier now make AudioSlides available with the journal articles.  I thought I’d try this out and see how it goes.  Obviously it can be a bit odd to make a recording of your voice available and I don’t really see that this would increase the number of people that read your article but thought it was worthwhile to try anyway.

http://audioslides.elsevier.com/viewerlarge.aspx?doi=10.1016/j.knosys.2013.07.002&resumeTime=NaN&resumeSlideIndex=0

Use and perceptions of worked example videos for first-year students studying mathematics in a primary education degree

Conference paper accepted for upcoming DELTA in Kiama, NSW (Australia).  I really got a lot out of DELTA in Rotorua, NZ – so I’ll be looking forward to attending this conference.  This particular project was undertaken with the help of some 3rd year science students.

Title: Use and perceptions of worked example videos for first-year students studying mathematics in a primary education degree

Authors: S. James, J. Brown, T. Gilbee and C. Rees

Abstract

Worked example videos have great potential to be useful for students when learning mathematics as they can work through the questions at their own pace, pausing as needed, but still learn from the way the demonstrator thinks and solves problems. We created worked example videos each week for a mathematics subject taught in the first year of a primary education degree and investigated student perceptions and their usage patterns.  An additional aspect of this undertaking was the inclusion of subtitles to make the videos accessible to hearing impaired and ESL students.  This report will reflect on the process of creating these videos, as well as some initial findings on their success.

 

Consensus measures constructed from aggregation functions and fuzzy implications

Recently accepted article in Knowledge Based Systems for a special issue on Consensus (edited by Enrique Herrera-Viedma). This extends more formally the measure presented at IFSA.  We look at the properties of a measure that should model the concept of consensus for a set of real inputs.  Although this work was mainly motivated by a connection between the form of the Bonferroni mean and the double-fuzzy integrals, I can also see some similarities between properties desired to model Consensus and those to model Evenness in ecology.

Title: Consensus measures constructed from aggregation functions and fuzzy implications

Authors: G. Beliakov, T. Calvo and S. James

Abstract

We focus on the problem of constructing functions that are able to measure the degree of consensus for a set of inputs provided over the unit interval.  When making evaluations based on inputs from multiple criteria, sources or experts, the resulting output can be seen as the value which best represents the individual contributions.   However it may also be desirable to know the extent to which the inputs agree.  Does the representative value reflect a universal opinion?  Or has there been a high degree of tradeoff?  We consider the properties relating to such consensus measures and propose two general models built component-wise from aggregation functions and fuzzy implications.

A generalization of the Bonferroni mean based on partitions

Paper recently presented at FUZZIEEE in Hyderabad, India.  This was a joint work with Radko Mesiar and continues our work on generalizing the Bonferroni mean, facilitating the modeling of mandatory requirements.

Title: A generalization of the Bonferroni mean based on partitions

Authors: G. Beliakov, S. James and R. Mesiar

Abstract:

The mean defined by Bonferroni in 1950 (known by the same name) averages all non-identical product pairs of the inputs.  Its generalizations to date have been able to capture unique behavior that may be desired in some decision-making contexts such as the ability to model mandatory requirements.  In this paper, we propose a composition that averages conjunctions between the respective means of a designated subset-size partition. We investigate the behavior of such a function and note the relationship within a given family as the subset size is changed. We found that the proposed function is able to more intuitively handle multiple mandatory requirements or mandatory input sets.

 

Two papers accepted for IFSA World Congress 2013

The papers below have both been accepted for focus sessions at the International Fuzzy Systems Association World Congress to be held in Edmonton, Alberta, Canada.

Title: Aggregating fuzzy implications to measure group consensus

Authors: G. Beliakov, T. Calvo and S. James

Abstract:

We approach the problem of measuring consensus for a set of real inputs by aggregating the fuzzy implication degrees between each pair of inputs.  We compare our operator with existing consensus measures in terms of their satisfaction of desirable properties.  The appeal of such an approach lies in the interpretability and flexibility that results from component-wise construction which we modeled on the Bonferroni mean.  We also outline some intentions for future research.

 

Title: Learning aggregation weights from 3-tuple comparison sets

Authors: G. Beliakov, S. James and D. Nimmo

Abstract:

An important task in multiple-criteria decision making is how to learn the weights and parameters of an aggregation function from empirical data.  We consider this in the context of quantifying ecological diversity, where such data is to be obtained as a set of pairwise comparisons specifying that one community should be considered more diverse than another.  A problem that arises is how to collect a sufficient amount of data for reliable model determination without overloading individuals with the number of comparisons they need to make.  After providing an algorithm for determining criteria weights and an overall ranking from such information, we then investigate the improvement in accuracy if ranked 3-tuples are supplied instead of pairs.  We found that aggregation models could be determined accurately from significantly fewer 3-tuple comparisons than pairs.

 

The R code relating to these papers will be made available soon.