Indexing the Environmental Quality Performance Based on A Fuzzy Inference Approach

International Conference on Information Technology and Digital Applications 2017

计算机科学

Iswari, Lizda^1

Department of Informatics, Universitas Islam Indonesia, Jalan Kaliurang Km. 14, Ngaglik Sleman Yogyakarta Indonesia, Indonesia^1

关键词: Environmental conditions;    Environmental performance;    Environmental quality indices;    Land cover;    Mamdani;    Mamdani fuzzy inferences;    Spatial;    Water index;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/325/1/012009/pdf
DOI  :  10.1088/1757-899X/325/1/012009

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

Environmental performance strongly deals with the quality of human life. In Indonesia, this performance is quantified through Environmental Quality Index (EQI) which consists of three indicators, i.e. river quality index, air quality index, and coverage of land cover. The current of this instrument data processing was done by averaging and weighting each index to represent the EQI at the provincial level. However, we found EQI interpretations that may contain some uncertainties and have a range of circumstances possibly less appropriate if processed under a common statistical approach. In this research, we aim to manage the indicators of EQI with a more intuitive computation technique and make some inferences related to the environmental performance in 33 provinces in Indonesia. Research was conducted in three stages of Mamdani Fuzzy Inference System (MAFIS), i.e. fuzzification, data inference, and defuzzification. Data input consists of 10 environmental parameters and the output is an index of Environmental Quality Performance (EQP). Research was applied to the environmental condition data set in 2015 and quantified the results into the scale of 0 to 100, i.e. 10 provinces at good performance with the EQP above 80 dominated by provinces in eastern part of Indonesia, 22 provinces with the EQP between 80 to 50, and one province in Java Island with the EQP below 20. This research shows that environmental quality performance can be quantified without eliminating the natures of the data set and simultaneously is able to show the environment behavior along with its spatial pattern distribution.