Analysis of an evolutionary algorithm for complex fuzzy cognitive map learning based on graph theory metrics and output concepts

Poczęta, Katarzyna; Kubuś, Łukasz; Yastrebov, Alexander

doi:10.1016/j.biosystems.2019.02.010

Cited by 23 publications

(10 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Indeed, results on noisy datasets show that changes in individual trajectories can be too abrupt for the FCMs, which are not currently able to cope. Although the high sensitivity of FCMs to noise was considered an asset in the past to identify 'corrupted' models [Taber, 1991], the lack of robustness with regard to noise has recently been considered as a problem [Orang et al, 2022]. Solutions have emerged for sparse FCMs [Wu and Liu, 2016], but developing them to improve the robustness of evolutionary learning algorithms while maintaining scalability remains an open problem and hence a core objective for future work.…”

Section: Discussionmentioning

confidence: 99%

“…Noteworthy complementary studies have used further constraints to guide the GA, for example to produce maps with desired structural properties such as sparsity [Stach et al, 2012, Poczeta et al, 2019. The use of GAs for FCMs remains an active research area, as illustrated by emerging applications of GAs to extended types of FCMs.…”

Section: Evolutionary Algorithmsmentioning

confidence: 99%

“…Our previous work showed that it was possible to create individual FCMs from data [Wozniak et al, 2022], but scaling to a large population was unfeasible. Although several studies have investigated time series in relation with FCMs, scalability to the number of data points has received little attention and many methods came with limitations such as "time series that are linear and tend to be trend stationary" [Salmeron and Froelich, 2016] or only predicting the next data point [Feng et al, 2021]. There is thus a need for methods that can flexibly follow the mental model of an individual over time (which may have any trajectory) and cope with a large number of measurements (which is increasingly needed as sensors are now feeding data to FCMs [Vaščák et al, 2021]).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Data Annotation for Support Ticket Data: A Literature Review

Fuchs,

Schnellbach,

Schmidt

et al. 2024

Proceedings of the Annual Hawaii International Conference on System Sciences

View full text Add to dashboard Cite

Models that capture the heterogeneous perspectives of individuals are essential to test tailored interventions, such as behavior change interventions. Although Fuzzy Cognitive Maps (FCMs) have a rich history in depicting systems, they were either developed at an individual level through facilitated sessions, or created for an entire population through machine learning. The need to automatically create individual FCMs from data has started to be addressed, but the proposed solution was computationally prohibitive and thus could not be deployed over a large population. In this work, we use a state-of-the-art evolutionary algorithm (CMA-ES) to create individual FCMs by leveraging the growing availability of longitudinal data. We demonstrate on a real-world case study that our solution is both accurate and fast to compute. Our experiments on synthetic data also show that our approach can scale to a large number of measurements, but it cannot currently be applied to highly noisy datasets.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Evolutionary Algorithmsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Data Annotation for Support Ticket Data: A Literature Review

Fuchs,

Schnellbach,

Schmidt

et al. 2024

Proceedings of the Annual Hawaii International Conference on System Sciences

View full text Add to dashboard Cite

show abstract

“…The FCM model's efficiency can be raised, results can be better understood, and the risk of overfitting can be decreased by lowering the number of concepts in the model. Numerous studies have demonstrated this increase in efficiency, three of which are [28,30,101]. Reduced models retain the model's predictive power while being more dependable, thorough, and efficient.…”

Section: Reduction Of the Fcm Via Experts' Knowledgementioning

confidence: 99%

A Fuzzy Cognitive Map and PESTEL-Based Approach to Mitigate CO2 Urban Mobility: The Case of Larissa, Greece

Kokkinos

Nathanail

2023

Sustainability

View full text Add to dashboard Cite

The CO2 reduction promise must be widely adopted if governments are to decrease future emissions and alter the trajectory of urban mobility. However, from a long-term perspective, the strategic vision of CO2 mitigation is driven by inherent uncertainty and unanticipated volatility. As these issues emerge, they have a considerable impact on the future trends produced by a number of exogenous and endogenous factors, including Political, Economic, Social, Technological, Environmental, and Legal aspects (PESTEL). This study’s goal is to identify, categorize, and analyze major PESTEL factors that have an impact on the dynamics of urban mobility in a rapidly changing environment. For the example scenario of the city of Larissa, Greece, a Fuzzy Cognitive Map (FCM) approach was employed to examine the dynamic interactions and behaviors of the connected criteria from the previous PESTEL categories. An integrative strategy that evaluates the interaction of linguistic evaluations in the FCM is used to include all stakeholders in the creation of a Decision Support System (DSS). The methodology eliminates the uncertainty brought on by a dearth of quantitative data. The scenarios in the study strands highlight how urbanization’s effects on sustainable urban transportation and the emergence of urban PESTEL actors impact on CO2 reduction decision-making. We focus on the use case of Larissa, Greece (the city of the CIVITAS program), which began putting its sustainable urban development plan into practice in 2015. The proposed decision-making tool uses analytics and optimization algorithms to point responsible authorities and decision-makers in the direction of Larissa’s sustainable urban mobility and eventually the decarbonization of the urban and suburban regions.

show abstract

“…Traditionally, domain experts allocate initial weights. Besides, Hebbian-based learning algorithms [41,42] and evolutionary algorithms [43][44][45] can improve the accuracy of the weights, and these new methods can effectively coordinate conflicts among experts. The obtained FCM model can be used to study the characteristics of complex systems.…”

Section: Introductionmentioning

confidence: 99%

Causal Inference of Optimal Control Water Level and Inflow in Reservoir Optimal Operation Using Fuzzy Cognitive Map

Liu

Zhou

et al. 2019

Water

View full text Add to dashboard Cite

Reservoir optimal operation (ROO) has always been a hot issue in the field of water resources management. Analysis of the relationship of optimal control water level and inflow is conducive to understanding and solving ROO under deterministic inflow conditions. The current research uses a fuzzy cognitive map (FCM) as a tool to effectively model complex systems and then extracts systematic relationship diagrams from the dataset. A new fuzzy cognitive map with offset (FCM-O) is proposed to overcome the causal inference error caused by non-linear mapping of the activation function in a traditional FCM. With the application of inferring the causal relationship between the optimal control water level and inflow of ROO for the Three Gorges Reservoir (TGR), the experimental results show that, compared with FCM in the min data error, FCM-O reduces 11.11% and 7.14% in the training and the testing, respectively. Also, the experimental results of FCM-O are more reasonable than those of FCM. Finally, the following conclusions about the causal inference of optimal control water level and inflow in ROO for TGR are drawn: (1) The optimal control water level in September, October and November needs to be raised as much as possible to raise the water head of power generation, which is mainly affected by the constraints of the maximum operating water level of the reservoir rather than inflow; (2) the optimal control water level in January, February and March is positively affected by the inflow of the adjacent months; (3) the optimal control water level in April is due to the approaching flood season. In order to prevent water discarding, the water level is low and the optimum operation space is small. All of those shows that FCM-O is more competent than FCM in the causal relationship between optimal control water level and inflow in ROO.

show abstract

Analysis of an evolutionary algorithm for complex fuzzy cognitive map learning based on graph theory metrics and output concepts

Cited by 23 publications

References 22 publications

Data Annotation for Support Ticket Data: A Literature Review

Data Annotation for Support Ticket Data: A Literature Review

A Fuzzy Cognitive Map and PESTEL-Based Approach to Mitigate CO2 Urban Mobility: The Case of Larissa, Greece

Causal Inference of Optimal Control Water Level and Inflow in Reservoir Optimal Operation Using Fuzzy Cognitive Map

Contact Info

Product

Resources

About