Ayahiko Niimi

Genetic programming combined with association rule algorithm for decision tree construction

Genetic programming (GP) usually has a wide search space and a high flexibility. So, GP may search for a global optimum solution. In general, GPs learning speed is not so fast. The Apriori Algorithm is one of the association rule algorithms. It can be applied to large databases, but it is difficult to define its parameters without experience. We propose a rule generation technique from a database using GP combined with an association rule algorithm. It takes rules generated by the association rule algorithm as the initial individual of GP. The learning speed of GP is improved by the combined algorithm. To verify the effectiveness of the proposed method, we apply it to the decision tree construction problem from the UCI Machine Learning Repository. We compare the result of proposed method with prior ones.

Data Mining for Distributed Databases with Multiagents

We propose a technique for using multiagent technology in data mining intended for two or more text databases. In this paper, we discuss data mining method based on text (text mining), but our proposed method is not a method of specializing in text mining. First of all, we introduce some typical techniques as a technique of data mining to text database. Next, multiagent technology is described. We propose data mining technique using multiagent technology. The proposed technique is applied to document databases, and discuss its results.

Rule Discovery Technique Using Genetic Programming Combined with Apriori Algorithm

Various techniques have been proposed for rule discovery using classification learning. In general,the learning speed of a system using genetic programming (GP) [1] is slow. However,a learning system which can acquire higher-order knowledge by adjusting to the environment can be constructed,b ecause the structure is treated at the same time.

Extension of Multiagent Data Mining for Distributed Databases

We proposed a technique for using multiagent technology in data mining intended for two or more text databases. In this paper, we discuss data mining method based on text (text mining), but our proposed method is not a method of specializing in text mining. First, we proposed data mining technique using multiagent technology. The proposed technique is applied to document databases, and discuss its results. Next, we extend the proposed technique with Stem algorithm, English morphological analysis, changed development language, adding the experiment data, and adding data mining algorithm.

Extended genetic programming using reinforcement learning operation

Genetic programming (GP) usually has a wide search space and a high flexibility, so GP may search for a global optimum solution. But GP has two problems. One is slow learning speed and a huge number of generations spending. The other is difficulty in operating continuous numbers. GP searches many tree patterns including useless node trees and meaningless expression trees. In general, GP has three genetic operators (mutation, crossover and reproduction). We propose an extended GP learning method including two new genetic operators, pruning (pruning redundant patterns) and fitting (fitting random continuous nodes). These operators have a reinforcement learning effect, and improve the efficiency of GPs search. To verify the validity of the proposed method, we developed a medical diagnostic system for the occurrence of hypertension. We compared the results of the proposed method with prior ones.

Construction of school temperature measurement system with sensor network

We propose the sensor network system using the microcomputer board that can connect to the Internet. This proposed system can acquire information from the sensor of the microcomputer group arranged on the network, and can view collected information on Web browser. In this paper, it is shown to be able to construct easily the microcomputers sensor network which is combined microcomputer modules (Micro Cube) and the database server and the Web application server. The system that measured the room temperature in school campus was constructed, it has run for four months, and the effectiveness is verified.

Combined method of genetic programming and association rule algorithm

Genetic programming (GP) usually has a wide search space and a high flexibility. Therefore, GP may search for global optimum solution. But, in general, GPs learning speed is not so fast. An apriori algorithm is one of association rule algorithms. It can be applied to a large database. But it is difficult to define its parameters without experience. We propose a rule generation technique from a database using GP combined with an association rule algorithm. It takes rules generated by the association rule algorithm as initial individual of GP. The learning speed of GP is improved by the combined algorithm. To verify the effectiveness of the proposed method, we apply it to the decision tree construction problem from the University of California at Irvine (UCI) machine-learning repository, and rule discovery problem from the occurrence of the hypertension database. We compare the results of the proposed method with prior ones.

Deep learning for credit card data analysis

In this paper, two major applications are introduced to develop advanced deep learning methods for credit-card data analysis. The proposed methods are validated using benchmark experiments with other machine learnings. The experiments confirm that deep learning exhibits similar accuracy to the Gaussian kernel SVM.

Extended Genetic Programming Using Apriori Algorithm for Rule Discovery

Genetic programming (GP) usually has a wide search space and can use tree structure as its chromosome expression. So, GP may search for global optimum solution. But, in general, GPs learning speed is not so fast. Apriori algorithm is one of algorithms for generation of association rules. It can be applied to large database. But, It is difficult to define its parameters without experience. We propose a rule discovery technique from a database using GP combined with association rule algorithm. It takes rules generated by the association rule algorithm as initial individual of GP. The learning speed of GP is improved by the combined algorithm. To verify the effectiveness of the proposed method, we apply it to the meningoencephalitis diagnosis activity data in a hospital. We got domain experts comments on our results. We discuss the result of proposed method with prior ones.

Object oriented approach to combined learning of decision tree and ADF GP

There are many learning methods for classification systems. Genetic programming (one of the methods) can change trees dynamically, but its learning speed is slow. Decision tree methods using C4.5 construct trees quickly, but the network may not classify correctly when the training data contains noise. For such problems, we proposed an object oriented approach, and a learning method that combines decision tree making method (C4.5) and genetic programming. To verify the validity of the proposed method we developed two different medical diagnostic systems. One is a medical diagnostic system for the occurrence of hypertension the other is for the meningoencephalitis. We compared the results of proposed method with prior ones.