Jesada Kajornrit

Estimation of missing precipitation records using modular artificial neural networks

Estimation of missing precipitation records is one of the important tasks in hydrological study. The completeness of precipitation data leads to more accurate results from the hydrological models. This study proposes the use of modular artificial neural networks to estimate missing monthly rainfall data in the northeast region of Thailand. The simultaneous rainfall data from neighboring control stations are used to estimate missing rainfall data at the target station. The proposed method uses two artificial neural networks to learn the generalized relationship of rainfall recorded in dry and wet periods. Inverse distance weighting method and optimized weight of subspace reconstruction method are used to aggregate the final estimation value from both networks. The experimental results showed that modular artificial neural networks provided a higher accuracy than single artificial neural network and other conventional methods in terms of mean absolute error.

Rainfall prediction in the northeast region of Thailand using Modular Fuzzy Inference System

In water management systems, accurate rainfall forecasting is indispensable for operation and management of reservoir, and flooding prevention because it can provide an extension of lead-time of the flow forecasting. In general, time series prediction has been widely applied to predict rainfall data. The conventional time series prediction models or artificial neural networks can be used to perform this task. However, such models are difficult to interpret by human analyst. From a hydrologists point of view, the accuracy of the prediction and understanding the prediction model are equally important. This study proposes the use of a Modular Fuzzy Inference System (Mod FIS) to predict monthly rainfall data in the northeast region of Thailand. The experimental results show that the proposed model can be a good alternative method to provide both accurate results and human-understandable prediction mechanism.

Cluster validation methods for localization of spatial rainfall data in the northeast region of Thailand

In order to develop spatial interpolation models for a large area, localization is required to partition global spatial data into local areas. Fuzzy c-means clustering are normally used to perform this task. However, the method needs prior information to determine the most proper number of cluster. This study proposes the use of two cluster validation methods, statistic-based method and simulation-based method to determine the optimal number of cluster for spatial data. The statistic-based method analyzes standard deviation of spatial data to determine the number of cluster, whereas the simulation-based method analyzes the training performance of artificial neural network. The proposed methods were applied to the spatial rainfall data in the northeast region of Thailand. The experimental results demonstrated that the proposed methods could provide reasonable results. The statistic-based method is statistically explainable for human analysts, whereas the simulation-based is an easy-to-use technique for cluster validation.

An integrated intelligent technique for monthly rainfall time series prediction

This paper proposes a methodology to create an interpretable fuzzy model for monthly rainfall time series prediction. The proposed methodology incorporates the advantages of artificial neural network, fuzzy logic and genetic algorithm. In the first step, the differences between the time series data are calculated and they are used to define the interval between the membership functions of a Mamdani-type fuzzy inference system. Next, artificial neural network is used to develop the model from input-output data and the established model is then used to extract the fuzzy rules. The parameters of the created fuzzy model are then optimized by using genetic algorithm. The proposed model was applied to eight monthly rainfall time series data in the northeast region of Thailand. The experimental results showed that the proposed model provided satisfactory prediction accuracy when compared to other commonly-used prediction models. Due to the interpretability nature of the model, human analysts can gain insight knowledge of the data to be modeled.

Improving Performance of Decision Trees for Recommendation Systems by Features Grouping Method

Recently, recommendation systems have become an important tool to support and improve decision making for educational purposes. However, developing recommendation systems is far from trivial and there are specific issues associated with individual problems. Low-correlated input features is a problem that influences the overall accuracy of decision tree models. Weak relationship between input features can cause decision trees work inefficiently. This paper reports the use of features grouping method to improve the classification accuracy of decision trees. Such method groups related input features together based on their ontologies. The new inherited features are then used instead as new features to the decision trees. The proposed method was tested with five decision tree models. The dataset used in this study were collected from schools in Nakhonratchasima province, Thailand. The experimental results indicated that the proposed method can improve the classification accuracy of all decision tree models. Furthermore, such method can significantly decrease the computational time in the training period.

An Integrated Intelligent Technique for Monthly Rainfall Spatial Interpolation in the Northeast Region of Thailand

Spatial interpolation is a method to create spatial continuous surface from observed data points. Spatial interpolation is important to water management and planning because it could provide estimation of rainfall at unobserved area. This paper proposes a methodology to analyze and establish an integrated intelligent spatial interpolation model for monthly rainfall data. The proposed methodology starts with determining the optimal number of sub-regions by means of standard deviation analysis and artificial neural networks. Once the optimal number of sub-regions is determined, a Mamdani fuzzy inference system is generated by fuzzy c-means and then optimized by genetic algorithm. Four case studies were used to evaluate the accuracy of the established models and compared with trend surface analysis and artificial neural networks. The experimental results demonstrated that the proposed methodology provided reasonable interpolation accuracy and the methodology gave human understandable fuzzy rules to human analysts.

An interpretable fuzzy monthly rainfall spatial interpolation system for the construction of aerial rainfall maps

Spatial interpolation methods are normally used to create aerial rainfall maps from remote measuring data collected by raingauge network. However, most spatial interpolation methods are not in the form of interpretable data models. This could make further analysis on the spatial data difficult. This paper proposes a methodology to analyze and establish an interpretable fuzzy model for monthly rainfall spatial interpolation. The proposed methodology integrates the benefits of various soft computing techniques. The final outcome is the proposal of an interpretable fuzzy model that allows human analysts to gain insight into the spatial data to be modeled. The accuracy of the model is evaluated by eight monthly rainfall data in the northeast region of Thailand. The interpretability of the model is assessed by the interpretable fuzzy modeling criteria. The experimental results showed that the proposed methodology could be an alternative technique to create rainfall maps and to understand the characteristics of the spatial data.

A Modular Spatial Interpolation Technique for Monthly Rainfall Prediction in the Northeast Region of Thailand

Monthly rainfall spatial interpolation is an important task in hydrological study to comprehensively observe the spatial distribution of the monthly rainfall variable in the study area. A number of spatial interpolation methods have been successfully applied to perform this task. However, those methods mainly aim at achieving satisfactory interpolation accuracy and they disregard the interpolation interpretability. Without interpretability, human analysts will not be able to gain insight of the model of the spatial data. This paper proposes an alternative approach to achieve both accuracy and interpretability of the monthly rainfall spatial interpolation solution. A combination of fuzzy clustering, fuzzy inference system, genetic algorithm and modular technique has been used. The accuracy of the proposed method has been compared to the most commonly-used methods in geographic information systems as well as previously proposed method. The experimental results showed that the proposed model provided satisfactory interpolation accuracy in comparison with other methods. Besides, the interpretability of the proposed model could be achieved in both global and local perspectives. Human analysts may therefore understand the model from the derived model’s parameters and fuzzy rules.