Abstract | ||
---|---|---|
Existing approaches to time series prediction using fuzzy techniques do not consider time periods in a human understandable format. The current time series based fuzzy approaches lead to large feature spaces where the individual features could be prone to noise in the individual observations. This paper presents a Time Series based Interval Type 2 Fuzzy Logic System (TS-IT2FLS) where the time periods are considered in a fuzzy manner with easily understandable linguistic labels. We have performed several experiments using various data sets where the proposed TS-IT2FLS achieved 29% higher ROC-AUC score compared to other non deep-learning approaches while our suggested approach is more explainable compared to existing approaches. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1109/FUZZ-IEEE55066.2022.9882556 | 2022 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) |
Keywords | DocType | ISSN |
Time Series,Fuzzy,Explainable,XAI,FRBS | Conference | 1544-5615 |
ISBN | Citations | PageRank |
978-1-6654-6711-7 | 0 | 0.34 |
References | Authors | |
13 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ashish Bhatia | 1 | 0 | 0.34 |
Hani Hagras | 2 | 1747 | 129.26 |