Thumbnail Image

Theses

Browse

20171
Reset filters
Subject: Mathematics × Author: Bisht, Kamlesh × End date: 2017 × Start date: 2010 ×
Reset filters

Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    ThesisItemOpen Access
    Intuitionistic fuzzy and hesitant fuzzy sets based time series forecasting methods
    (G.B. Pant University of Agriculture and Technology, Pantnagar - 263145 (Uttarakhand), 2017-12) Bisht, Kamlesh; Sanjay Kumar
    In the present research work, various time series forecasting methods using intuitionistic fuzzy, hesitant fuzzy and dual hesitant fuzzy sets are developed to address issue of non stochastic uncertainty and non-stochastic hesitation. These methods are presented in form of various models. Model [1] and Model [4] are computational algorithm based higher order fuzzy time series forecasting models which are based on intuitionistic fuzzy and hesitant fuzzy respectively. Model [2] and Model [3] use hesitant fuzzy set to handle hesitancy in fuzzy time series forecasting. In the present research work, methodology of hesitant fuzzy time series forecasting method is also developed and presented in Model [5]. Model [6] is intuitionistic fuzzy time series forecasting model which is based on dual hesitant fuzzy set. Model [2], Model [3], Model [5] and Model [6] use max-min composition operation. Performance of these developed models over existing fuzzy time series and intuitionistic fuzzy time series forecasting methods is verified by implementing them on two time series data of University of Alabama and SBI share price. The outperformance of all models is tested using RMSE and AFER error measures. Validity of the developed models id tested using correlation coefficients, tracking signal, evaluation parameter and performance parameters. Superiority of developed models is also tested using two tailed t-test at the confidence level 1% and 5%. In forecasting enrollments of University of Alabama, the performance of Model [4] in terms of both RMSE and AFER is found better than model [1], Model [2], Model [3], Model [5] and Model [6]. In forecasting share prices of SBI, the performance of Model [6] in terms of both RMSE and AFER is found better than Model [1], Model [2], Model [3], Model [4] and Model [5]. In forecasting enrollments of University of Alabama, Model [1] outperforms over Model [2] but its performance is equally good as of Model [3], Model [4], Model [5] and Model [6]. Model [2] outperforms over Model [4] and Model [5] but its performance is good as Model [3] and Model [6]. Model [3] performs equally good as Model [4], Model [5] and Model [6]. Model [4] outperforms over Model [6] but its performance is good as Model [5]. Model [5] is equally good in performance as the Model [6]. In forecasting SBI share prices, Model [1] outperforms over Model [2], Model [3] but its performance is good as Model [4], Model [5]. Model [2] outperforms over Model [4], Model [5] and Model [6] but its performance is good as Model [3]. Model [3] outperforms over Model [4], Model [5] and Model [6]. Model [4] performed equally good as Model [5] and Model [6]. Model [5] outperforms over Model [6]. All developed models not only handle the non-stochastic uncertainty and hesitation but also enhance the accuracy in forecasted enrollments and financial time series data of SBI share price.