Towards fuzzy linguistic Markov chains

Villacorta, Pablo J.; Verdegay, José L.; Pelta, David A.

doi:10.2991/eusflat.2013.106

Cited by 4 publications

(6 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here,Ã ĚB Ø µÃpxq ě µBpxq @x P R, the symbol ' stands for the fuzzy sum, and 1 χ is the real number 1 represented as a fuzzy number (singleton). Note this is a necessary condition to ensure that Dompαq is not empty, as proved in Villacorta, Verdegay, and Pelta (2013) and, therefore, our code first checks this condition and stops if it is not fulfilled.…”

Section: User-specified Fuzzy Transition Probabilitiesmentioning

confidence: 99%

FuzzyStatProb: AnRPackage for the Estimation of Fuzzy Stationary Probabilities from a Sequence of Observations of an Unknown Markov Chain

Villacorta¹,

Verdegay²

2016

J. Stat. Soft.

View full text Add to dashboard Cite

Markov chains are well-established probabilistic models of a wide variety of real systems that evolve along time. Countless examples of applications of Markov chains that successfully capture the probabilistic nature of real problems include areas as diverse as biology, medicine, social science, and engineering. One interesting feature which characterizes certain kinds of Markov chains is their stationary distribution, which stands for the global fraction of time the system spends in each state. The computation of the stationary distribution requires precise knowledge of the transition probabilities. When the only information available is a sequence of observations drawn from the system, such probabilities have to be estimated. Here we review an existing method to estimate fuzzy transition probabilities from observations and, with them, obtain the fuzzy stationary distribution of the resulting fuzzy Markov chain. The method also works when the user directly provides fuzzy transition probabilities. We provide an implementation in the R environment that is the first available to the community and serves as a proof of concept. We demonstrate the usefulness of our proposal with computational experiments on a toy problem, namely a time-homogeneous Markov chain that guides the randomized movement of an autonomous robot that patrols a small area.

show abstract

Section: User-specified Fuzzy Transition Probabilitiesmentioning

confidence: 99%

FuzzyStatProb: AnRPackage for the Estimation of Fuzzy Stationary Probabilities from a Sequence of Observations of an Unknown Markov Chain

Villacorta¹,

Verdegay²

2016

J. Stat. Soft.

View full text Add to dashboard Cite

show abstract

“…In our model we implement a similar approach to the one used by Villacorta et al [23], using five possible linguistic labels (see Figure 2) to evaluate the probability that the chain moves from one state to another. For the sake of simplification, trapezoidal membership functions (such as the ones suggested by Bonissone [12]) can be used, and the number of linguistic terms can also be adapted at will.…”

Section: Interaction Predictionmentioning

confidence: 99%

“…In a recent implementation of linguistic fuzzy Markov chains, Villacorta, Verdegay and Pelta [23] suggest to address the uncertainty of linguist judgements by introducing fuzzy probabilities as an additional layer that is placed on top of Buckley's fuzzy Markov chains. Given a discrete linguistic probability distribution, π 1 , · · · , π n , their sum must contain the singleton fuzzy number 1 χ , regardless of the type of fuzzy numbers involved or how the sum has been defined.…”

Section: Linguistic Fuzzy Markov Chainsmentioning

confidence: 99%

See 1 more Smart Citation

Towards a fuzzy-linguistic based social network sentiment-expression system

Perticone¹,

D'Aleo²,

Rizzo³

et al. 2015

Proceedings of the 2015 Conference of the International Fuzzy Systems Association and the European Society for Fuzzy Logic and

View full text Add to dashboard Cite

Liking allows users of Social Networks, blogs and online magazines to express their support of posts and artifacts by a simple click. Such function is very popular but lacks semantic power, and some platforms have augmented it by allowing to choose a pictographic depiction corresponding to a feeling. What is gained in depth is lost in simplicity, and the wide acceptance liking has enjoyed did not carried to the sentiment version. We outline a sentiment-expression hybrid system based on textual analysis and linguistic fuzzy Markov chains overcoming the intrinsic limitations of liking without burdening the user with complex choices.

show abstract

“…in application for speech recognition and image classification [31,20]. In a recent implementation of linguistic fuzzy Markov chains, Villacorta et al [42] suggest to address the uncertainty of linguistic judgements by introducing fuzzy probabilities as an additional layer that is placed on top of Buckley's fuzzy Markov chains. This approach is the one we have followed for our implementation.…”

Section: Our Approach To Fuzzy Text Analysismentioning

confidence: 99%

Towards the Improvement of Citizen Communication Through Computational Intelligence

Perticone

Tabacchi

2016

Studies in Systems, Decision and Control

View full text Add to dashboard Cite

When dealing with problems that arise from collective sharing of resources in metropolitan areas (i.e., energy, pollution, traffic, health) most of the interaction between citizens and local governance is usually carried out through the use of natural languages. Digital technologies allows smart cities residents to communicate with a broad range of experts (e.g. bureaucrats, legislators, urbanists, etc.) that routinely use technical terminology seldom accessible to the layperson, or linguistic styles that are not immediately understandable. Although information technology should encourage citizen participation in governance at many levels, the different levels of knowledge possessed by the actors can lead to incomprehension, as well as social exclusion. Computational Intelligence approaches can be used in order to alleviate such difficulties and improve the efficiency of communication through automation and collective cognitive systems. In this paper we discuss how use of techniques such as CWW and Fuzzy Classifiers can be beneficial toward the reduction of the communication gap between citizens and government.

show abstract

Towards fuzzy linguistic Markov chains

Cited by 4 publications

References 25 publications

FuzzyStatProb: AnRPackage for the Estimation of Fuzzy Stationary Probabilities from a Sequence of Observations of an Unknown Markov Chain

FuzzyStatProb: AnRPackage for the Estimation of Fuzzy Stationary Probabilities from a Sequence of Observations of an Unknown Markov Chain

Towards a fuzzy-linguistic based social network sentiment-expression system

Towards the Improvement of Citizen Communication Through Computational Intelligence

Contact Info

Product

Resources

About