Expert-Based and Computational Methods for Developing Fuzzy Cognitive Maps

Supporting underrepresented students in health sciences: a fuzzy cognitive mapping approach to program evaluation

BACKGROUND

The Students Training in Academia, Health, and Research (STAHR) Program at the University of Missouri-Kansas City (UMKC) strives to help students from low-income families that have experienced educational challenges due to poverty and prepare them to enter, persist, and graduate from a health sciences degree program at UMKC. Students in the program participated in fuzzy cognitive mapping (FCM) sessions to ensure that all voices of the program were heard to improve program implementation, and student success, and contribute to an equitable educational environment.

METHODS

Fuzzy Cognitive Mapping sessions for the 2020-2021 cohort of students (n = 52) were conducted to assess the strengths and weaknesses in program implementation, especially through the beginning of the COVID-19 pandemic. Students' maps were coded by a team of researchers and then confirmed using confirmatory factor analysis.

RESULTS

Statistical analyses reveal that mentorship, workshops, and social support helped students to work toward their goal of obtaining a professional health sciences degree, while a lack of time, remote learning, and outside stressors inhibited their opportunities for success.

CONCLUSIONS

The findings from a multipronged analysis of mapping data demonstrate the value of this innovative approach to the field, especially when looking to incorporate student voices.

Research directions in policy modeling: Insights from comparative analysis of recent projects

With the increased availability of data and the capacity to make sense of these data, computational approaches to analyze, model and simulate public policy evolved toward viable instruments to deliberate, plan, and evaluate them in different areas of application. Such examples include infrastructure, mobility, monetary, or austerity policies, policies on different aspects of societies (health, pandemic, skills, inclusion, etc.). Technological advances along with the evolution of theoretical models and frameworks open valuable opportunities, while at the same time, posing new challenges. The paper investigates the current state of research in the domain and aims at identifying the most pressing areas for future research. This is done through both literature research of policy modeling and the analysis of research and innovation projects that either focus on policy modeling or involve it as a significant component of the research design. In the paper, 16 recent projects involving the keyword policy modeling were analyzed. The majority of projects concern the application of policy modeling to a specific domain or area of interest, while several projects tackled the cross-cutting topics (risk and crisis management). The detailed analysis of the projects led to topics of future research in the domain of policy modeling. Most prominent future research topics in policy modeling include stakeholder involvement approaches, applicability of research results, handling complexity of models, integration of models from different modeling and simulation paradigms and approaches, visualization of simulation results, real-time data processing, and scalability. These aspects require further research to appropriately contribute to further advance the field.

Prediction of DNA-Binding Residues in Local Segments of Protein Sequences with Fuzzy Cognitive Maps

While protein-DNA interactions are crucial for a wide range of cellular functions, only a small fraction of these interactions was annotated to date. One solution to close this annotation gap is to employ computational methods that accurately predict protein-DNA interactions from widely available protein sequences. We present and empirically test first-of-its-kind predictor of DNA-binding residues in local segments of protein sequences that relies on the Fuzzy Cognitive Map (FCM) model. The FCM model uses information about putative solvent accessibility, evolutionary conservation, and relative propensities of amino acid to interact with DNA to generate putative DNA-binding residues. Empirical tests on a benchmark dataset reveal that the FCM model secures AUC = 0.72 and outperforms recently released hybridNAP predictor and several popular machine learning methods including Support Vector Machines, Naïve Bayes, and k-Nearest Neighbor. The improvements in the predictive performance result from an intrinsic feature of FCMs that incorporate relations between the input features, besides the relations between the inputs and output that are modelled by other algorithms. We also empirically demonstrate that use of a short sliding window results in further improvements in the predictive quality. The funDNApred webserver that implements the FCM predictor is available at http://biomine.cs.vcu.edu/servers/funDNApred/.

A FCM-Based Dynamic Modeling of ERP Implementation Critical Failure Factors

Implementation of Enterprise Resource Planning systems (ERPs) is a complex and costly process which usually results in serious failures. Numerous factors affect these projects implementation due to their size, complexity and high chance of failure. Therefore, identifying these factors in ERP projects is a critical issue. The majority of previous studies and research projects have been conducted in identifying ERP Critical Success Factors (CSFs) rather than Critical Failure Factors (CFFs). In order to help practitioners, this paper studies the CFFs in this kind of projects. Unfortunately, the implications of interdependency among failure factors are usually underestimated by project managers and decision makers since they are difficult to model and analyze. With this in mind, the authors have built Fuzzy Cognitive Maps (FCMs) of failure factors in ERP implementation projects. The main advantage of FCM lies in them being capable of modeling complex phenomena based on the experts' perceptions. This tool models uncertainty and related events, imitating human reasoning. Moreover, FCMs enable the developing of forecasting exercises through simulations. Practitioners would thus assess the joint influence of ERP implementation failure factors on project outcomes. The results make known to practitioners which problems will arise if the failure factors are not treated, and how these will impact on the outcomes of projects. Therefore, the tool proposed would help them to manage ERP implementation projects in a more effective and proactive way.

Learning of fuzzy cognitive maps using density estimate

Fuzzy cognitive maps (FCMs) are convenient and widely used architectures for modeling dynamic systems, which are characterized by a great deal of flexibility and adaptability. Several recent works in this area concern strategies for the development of FCMs. Although a few fully automated algorithms to learn these models from data have been introduced, the resulting FCMs are structurally considerably different than those developed by human experts. In particular, maps that were learned from data are much denser (with the density over 90% versus about 40% density of maps developed by humans). The sparseness of the maps is associated with their interpretability: the smaller the number of connections is, the higher is the transparency of the map. To this end, a novel learning approach, sparse real-coded genetic algorithms (SRCGAs), to learn FCMs is proposed. The method utilizes a density parameter to guide the learning toward a formation of maps of a certain predefined density. Comparative tests carried out for both synthetic and real-world data demonstrate that, given a suitable density estimate, the SRCGA method significantly outperforms other state-of-the-art learning methods. When the density estimate is unknown, the new method can be used in an automated fashion using a default value, and it is still able to produce models whose performance exceeds or is equal to the performance of the models generated by other methods.