Swati Aggarwal

Hybrid model for medical diagnosis using Neutrosophic Cognitive Maps with Genetic Algorithms

This paper presents a combination of two soft computing techniques: Neutrosophic cognitive maps and Genetic algorithm for modeling of medical disease diagnosis. Earlier, medical decision support system was proposed by many researchers where cognitive maps along with genetic algorithm were experimented. The hybrid model of fuzzy cognitive maps with genetic algorithm was implemented to handle situations where decisions are not clearly distinct. In real world situation data is not always consistent so, authors proposed a new hybrid model of Neutrosophic Cognitive Maps with Genetic Algorithms to handle indeterminacy. The proposed model will provide distinct diagnosis of disease in medical decision support system.

Neutrosophic Trust Evaluation Model in B2C E-Commerce

Trust is an important term in the context of E-Commerce. It lies at the top in the present trend of E-Commerce. Trust lists the diverse prospects of trustworthiness that exist between the merchant and customer, inducing a better customer liking and Business-to-Customer (B2C) E-Commerce. Considering the imprecise nature of E-Commerce trust, various researchers proposed different trust models and integrated them with fuzzy logic to handle inherent uncertainty. Conventional models of trust are based on subjective logic which falls short in mapping the real-time environment of E-commerce that deals with tentative behavioural values. Though fuzzy logic representation of the facts is a way to deal with improbability but it fails to capture the indeterminacy and false values given by respondents during survey. Authors in this chapter have attempted to target the indeterminacy involved while capturing the perception of respondents during survey for any website. Quite recently, neutrosophic logic (NL) has been proposed by Florentine Smarandache that gives a mathematical model for representing uncertainty, vagueness, ambiguity, imprecision, incompleteness, inconsistency, redundancy and contradictions. All the factors stated are very integral to human thinking, as it is very rare that we tend to conclude/judge in definite environments, imprecision of human systems could be due to the imperfection of knowledge that the human receives (observation) from the external world. Imperfection leads to a doubt about the value of a variable, a decision to be taken or a conclusion to be drawn for the actual system. This chapter suggests computation of perceived trust value by integrating a neutrosophic logic with the proposed fuzzy based trust model that considers all the chief features which affect the trust in E-Commerce.

Multi-attribute data classification using Neutrosophic probability

Fuzzy classification is very necessary because it has the ability to use interpretable rules. It has got control over the limitations of crisp rule based classification. This paper mainly deals with classification using fuzzy probability and Neutrosophic probability. Classification based on Neutrosophic probability employs Neutrosophic logic and Neutrosophic probability for its working and is compared with classification based on fuzzy probability on the basis of parameters such as probability and ambiguity in the results. Classification based on fuzzy and Neutrosophic probability are implemented on appendicitis dataset from Knowledge extraction based on evolutionary learning.

A Novel Approach for Data Classification Using Neutrosophic Entropy

Fuzzy classification is very necessary because it has the ability to use interpretable rules. It has got control over the limitations of crisp rule-based classification. This paper mainly deals with classification using fuzzy probability and Neutrosophic probability. Classification based on Neutrosophic probability employs Neutrosophic logic, Neutrosophic probability, and Neutrosophic entropy for its working and is compared with classification based on fuzzy probability on the basis of parameters such as probability and ambiguity in the results. Classification based on fuzzy and Neutrosophic probabilities is implemented on appendicitis dataset from knowledge extraction based on evolutionary learning.

Using fuzzy c means and multi layer perceptron for data imputation: Simple v/s complex dataset

Data imputation is the process of filling in the missing value to generate complete records. Complete databases can be analyzed more accurately in comparison to incomplete databases. This paper implements a 2-stage hybrid model for filling in the missing values. Also the effect of the proposed model over simple and complex dataset with varying percentage of missing value and varying value of fuzzifier is evaluated. The accuracy of the model is checked with Mean Absolute Percentage Error (MAPE). The result obtained shows that the proposed model is more accurate in filling multiple values in a record compared to stage 1 alone.

Cognitive maps analysis of a social evil: Corruption

These days corruption can be seen in every section of society. If certain changes are not made, corruption will continue to be an issue in the society. This paper surveys and discusses various issues related to the causes of corruption using soft computing technique known as FCM (Fuzzy Cognitive Mapping) and E-FCM (Extended Fuzzy Cognitive Maps). A comparative study between FCM and E-FCM is conducted showing E-FCM is a better technique than FCM.

Understanding India's skewed sex ratio problem using cognitive maps

Recent studies have confirmed that the problem of female deficit in India is the result of sex selective abortion, negligence of girls, infanticides and foeticide, preference of son for the preservation for clan, gender bias and emergence of new technologies in medical field. This paper addresses this critical problem of gender inequality in India using fuzzy cognitive maps. Fuzzy cognitive maps model the world as a collection of classes and causal relation between classes. Authors have attempted to explore the relationships and the relative impacts among the causes; that fuels this crucial problem.

Experimenting with neutrosophic ontologies for medical data classification

Ontology helps us to think about the real world with its semantic constraints. But there exist different uncertainties and ambiguities that cannot be considered using traditional ontologies. This paper mainly deals with classification using fuzzy ontology and Neutrosophic ontology. Classification based on Neutrosophic ontology incorporates Neutrosophic logic for its working and is compared with classification based on fuzzy ontology on appendicitis dataset from Knowledge extraction based on evolutionary learning.