Jung Yi Lin

Classifier design with feature selection and feature extraction using layered genetic programming

This paper proposes a novel method called FLGP to construct a classifier device of capability in feature selection and feature extraction. FLGP is developed with layered genetic programming that is a kind of the multiple-population genetic programming. Populations advance to an optimal discriminant function to divide data into two classes. Two methods of feature selection are proposed. New features extracted by certain layer are used to be the training set of next layers populations. Experiments on several well-known datasets are made to demonstrate performance of FLGP.

Learning discriminant functions with fuzzy attributes for classification using genetic programming

Abstract Classification is one of the important tasks in developing expert systems. Most of the previous approaches for classification problem are based on classification rules generated by decision trees. In this paper, we propose a new learning approach based on genetic programming to generate discriminant functions for classifying data. An adaptable incremental learning strategy and a distance-based fitness function are developed to improve the efficiency of genetic programming-based learning process. We first transform attributes of objects into fuzzy attributes and then a set of discriminant functions is generated based on the proposed learning procedure. The set of derived functions with fuzzy attributes gives high accuracy of classification and presents a linear form. Hence, the functions can be transformed into inference rules easily and we can use the rules to provide the building of rule base in an expert system.

Designing a classifier by a layered multi-population genetic programming approach

This paper proposes a method called layered genetic programming (LAGEP) to construct a classifier based on multi-population genetic programming (MGP). LAGEP employs layer architecture to arrange multiple populations. A layer is composed of a number of populations. The results of populations are discriminant functions. These functions transform the training set to construct a new training set. The successive layer uses the new training set to obtain better discriminant functions. Moreover, because the functions generated by each layer will be composed to a long discriminant function, which is the result of LAGEP, every layer can evolve with short individuals. For each population, we propose an adaptive mutation rate tuning method to increase the mutation rate based on fitness values and remaining generations. Several experiments are conducted with different settings of LAGEP and several real-world medical problems. Experiment results show that LAGEP achieves comparable accuracy to single population GP in much less time.

Learning effective classifiers with Z-value measure based on genetic programming

This paper presents a learning scheme for data classification based on genetic programming. The proposed learning approach consists of an adaptive incremental learning strategy and distance-based fitness functions for generating the discriminant functions using genetic programming. To classify data using the discriminant functions effectively, the mechanism called Z-value measure is developed. Based on the Z-value measure, we give two classification algorithms to resolve ambiguity among the discriminant functions. The experiments show that the proposed approach has less training time than previous GP learning methods. The learned classifiers also have high accuracy of classification in comparison with the previous classifiers.

A Function-Based Classifier Learning Scheme Using Genetic Programming

Classification is an important research topic in knowledge discovery and data mining. Many different classifiers have been motivated and developed of late years. In this paper, we propose an effective scheme for learning multicategory classifiers based on genetic programming. For a k-class classification problem, a training strategy called adaptive incremental learning strategy and a new fitness function are used to generate k discriminant functions. We urge the discriminant functions to map the domains of training data into a specified interval, and thus data will be assigned into one of the classes by the values of functions. Furthermore, a Z-value measure is developed for resolving the conflicts. The experimental results show that the proposed GP-based classification learning approach is effective and performs a high accuracy of classification.

Applying layered multi-population genetic programming on learning to rank for information retrieval

Information retrieval (IR) returns a relative ranking of documents with respect to a user query. Learning to rank for information retrieval (LR4IR) employs supervised learning techniques to address this problem, and it aims to produce a ranking model automatically for defining a proper sequential order of related documents based on the query. The ranking model determines the relationship degree between documents and the query. In this paper an improved version of RankGP is proposed. It uses layered multi-population genetic programming to obtain a ranking function which consists of a set of IR evidences and particular predefined operators. The proposed method is capable to generate complex functions through evolving small populations. In this paper, LETOR 4.0 was used to evaluate the effectiveness of the proposed method and the results showed that the method is competitive with other LR4IR Algorithms.

Evolving GPS position correction function using genetic programming

Many mobile devices embeds Global positioning system (GPS) to enable tour navigation or path programming. Those low-cost consumer-grade GPS receivers usually do not have high accuracy so that a position correction algorithm is therefore necessary. This paper proposed a correction technique using genetic programming. This technique requires only two GPS receivers and a known position. Using position information gathered by the receivers we are capable of predicting the correct position. The proposed technique can be implemented without modifying hardware devices or settings. The algorithm generates a correction function constructed by features of NMEA (national Marine Electronics Association) sentences, which is a standard format and is common in most GPS receivers. Experiments are conducted to demonstrate performance of the proposed technique. Positioning error could be reduced significantly.

Position Correction on Consumer-Grade GPS Using Genetic Programming

Consumer-grade global positioning system (GPS) becomes common and popular in these days because of its low cost along with acceptable accuracy. A GPS receiver cannot always obtain precise position because it affected by errors from either satellites or the receiver itself. Many researchers proposed effective approaches to improve positioning accuracy of GPS receivers. In this paper, we propose a method based on the concept of differential correction using two consumer-grade GPS receivers and genetic programming (GP). The proposed method generates a correction function through GPS information gathered by GPS receivers and a known position. Any GPS receiver which transfers NMEA (National Marine Electronics Association) sentence information can be used for the proposed method. the method could be implemented on various GPS-embedded devices without modifying hardware components.

Fitness enhancement of layered architecture genetic programming

Layered architecture genetic programming (LAGEP) has been applied on variety classification problems. It organizes populations as layers. Populations in different layers evolve with different training sets. Individuals produced by populations of layer Li transform training instances into new ones. Populations in Li+1 then evolve with the new training set instead of evolve with the original given training set. Each population in Li produces one feature for the new training instances. New training instances could have fewer features and are easier to be classified. Such mechanism makes consecutive layer gain better fitness value than preceding layers do. At this paper, we intend to analyze the enhancement of fitness value over all layers. We conduct experiments with a high-dimensional gene expression dataset to show the fitness enhancement.

Cancer classification using microarray and layered architecture genetic programming

An important problem of cancer diagnosis and treatment is to distinguish tumors from malignant or benign. Classifying tumors correctly leads us to target specific therapies properly to maximizing efficiency and reducing toxicity. Through the microarray technology, it is possible that monitoring expression in cells for numerous of genes simultaneously. Therefore we are allowed to use potential information hidden in the gene expression data to build a more accurate and more reliable classification model on tumor samples. In this paper we intend to investigate a new approach for cancer classification using genetic programming and microarray gene expression profiles. The layered architecture genetic programming (LAGEP) is applied to build the classification model. Some typical cancer gene expression datasets are validated to demonstrate the classification accuracy of the proposed model.