Targets of Unequal Importance Using the Concept of Stratification in a Big Data Environment
- 189 Downloads
The concept of stratification (CST) has recently been proposed as an innovative approach in problem solving. CST takes a recursive approach to solve problems. It considers a system which has to transition through states until it arrives to a state which belongs to a desired set of states, namely a target set. The states can be stratified by enlarging the target (absorbing adjacent states). Incremental enlargement is a means to identify possible paths to achieve the target. Such an enlargement can also be used to degrade the target when the original target is not reachable. Although the characteristics of the concept, such as incremental enlargement, enhance its potential application in robotics, artificial intelligence, and planning and monitoring, there is a major shortcoming in the approach, namely its inability to consider targets of unequal importance. This study considers two targets of unequal importance for the system in CST, labelled Bi-Objective CST model (BOCST). In comparison with the original proposed CST model in this research, a version of CST with finite states which is much easier to apply than the original CST is proposed, labelled fuzzy CST. Following that, a combination of Fuzzy CST and BOCST (FBO-CST) is proposed. The model is then employed to address a restaurant selection problem using data from Google. The example illustrates how the model should be applied in a big data environment. By defining finite state CST and considering targets of unequal importance, this study is expected to facilitate future applications of CST.
KeywordsConcept of stratification (CST) Granulation Unequal targets Unequal conjunction Big data
The paper was inspired by the one of the last contributions of a world-renowned scientist, Professor Lotfi Zadeh. Zadeh’s theories of fuzzy logic have influenced many researchers in the last few decades, and we believe that CST also provides great potential for future research. We were shocked when we were informed recently that the world had lost one of its great researchers. May he rest in peace.
- 1.Zaniolo, C.: Expressing and supporting E ciently Greedy algorithms as locally stratified logic programs (2015)Google Scholar
- 3.Dascalu, S.M., et al.: Stratified programming integrated development environment (SPIDER). In: IASSE (2003)Google Scholar
- 6.Balmin, A., et al.: Stratified sampling using adaptive parallel data processing. Google Patents (2017)Google Scholar
- 7.Zhang, Y., et al.: Fast model predictive control for multilevel cascaded h-bridge statcom with polynomial computation time. IEEE Trans. Ind. Electron. 63(8), 5231–5243 (2016)Google Scholar
- 9.Asadabadi, M., Saberi, M., Chang, E.: Logistic informatics modelling using concept of stratification (CST). In: IEEE Fuzzy 2017. IEEE Xplore (2017)Google Scholar
- 16.Azadeh, A., et al.: Z-AHP: a Z-number extension of fuzzy analytical hierarchy process. In: 7th IEEE International Conference on Digital Ecosystems and Technologies (DEST). IEEE (2013)Google Scholar
- 17.Balal, E., Cheu, R.L.: Modeling of lane changing decisions: comparative evaluation of fuzzy inference system, support vector machine and multilayer feed-forward neural network (2017)Google Scholar
- 21.Google: Google Transparency Report (2017)Google Scholar