The Fuzzy Lattice Reasoning Classifier for mining environmental data
This chapter introduces a rule-based perspective on the framework of fuzzy lattices, and the Fuzzy Lattice Reasoning (FLR) classifier. The notion of fuzzy lattice rules is introduced, and a training algorithm for inducing a fuzzy lattice rule engine from data is specified. The role of positive valuation functions for specifying fuzzy lattices is underlined and non-linear (sigmoid) positive valuation functions are proposed, that is an additional novelty of the chapter. The capacities for learning of the FLR classifier using both linear and sigmoid functions are demonstrated in a real-world application domain, that of air quality assessment. To tackle common problems related to ambient air quality, a machine learning approach is demonstrated in two applications. The first one is for the prediction of the daily vegetation index, using a dataset from Athens, Greece. The second concerns with the estimation of quartely ozone concentration levels, using a dataset from Valencia, Spain.
I. N. Athanasiadis, The Fuzzy Lattice Reasoning Classifier for mining environmental data, Computational Intelligence Based on Lattice Theory, Studies in Computational Intelligence, pg. 175-193, 2007, Springer-Verlag, doi:10.1007/978-3-540-72687-6.
You might also enjoy (View all publications)
- Mixing process-based and data-driven approaches in yield prediction
- Combining telecom data with heterogeneous data sources for traffic and emission assessments - an agent-based approach
- A weakly supervised framework for high-resolution crop yield forecasts