Thomas M. Tirpak

Evolving accurate and compact classification rules with gene expression programming

Classification is one of the fundamental tasks of data mining. Most rule induction and decision tree algorithms perform a local, greedy search to generate classification rules that are often more complex than necessary. Evolutionary algorithms for pattern classification have recently received increased attention because they can perform global searches. In this paper, we propose a new approach for discovering classification rules by using gene expression programming (GEP), a new technique of genetic programming (GP) with linear representation. The antecedent of discovered rules may involve many different combinations of attributes. To guide the search process, we suggest a fitness function considering both the rule consistency gain and completeness. A multiclass classification problem is formulated as multiple two-class problems by using the one-against-all learning method. The covering strategy is applied to learn multiple rules if applicable for each class. Compact rule sets are subsequently evolved using a two-phase pruning method based on the minimum description length (MDL) principle and the integration theory. Our approach is also noise tolerant and able to deal with both numeric and nominal attributes. Experiments with several benchmark data sets have shown up to 20% improvement in validation accuracy, compared with C4.5 algorithms. Furthermore, the proposed GEP approach is more efficient and tends to generate shorter solutions compared with canonical tree-based GP classifiers.

Optimization of high-speed multistation SMT placement machines using evolutionary algorithms

Surface mount technology (SMT) is a robust methodology that has been widely used in the past decade to produce circuit boards. Analyses of the SMT assembly line have shown that the automated placement machine is often the bottleneck, regardless of the arrangement of these machines (parallel or sequential) in the assembly line. Improving and automating the placement machine is a key issue for increasing SMT production line throughput. This paper presents experimental results using genetic algorithms to optimize the feeder slot assignment problem for a high-speed parallel, multistation SMT placement machine. Four crossover operators, four selection methods, and two probability settings are used in our experiments. A penalty function is used to handle constraints. A comparison of genetic algorithms with several other optimization methods (human experts, vendor supplied software, expert systems, and local search) is presented, which supports the use of genetic algorithms for this problem.

Optimization of high-mix printed circuit card assembly using genetic algorithms

The purpose of this paper is to present an overview of the factors affecting the cycle time of printed circuit card assembly (PCCA) in high-mix environments and demonstrate a technique for improving machine throughput. We have concentrated our research on optimizing the portion of the PCCA manufacturing process performed by high-speed placement machines (chip shooters). A crucial factor affecting the throughput of a chip shooter is the assignment of components to the feeder slots. Genetic algorithms were employed to find a near optimal assignment of the feeder carriage. Results for various genetic operators in this problem domain are presented.

A visual data mining framework for convenient identification of useful knowledge

Data mining algorithms usually generate a large number of rules, which may not always be useful to human users. In this project, we propose a novel visual data-mining framework, called Opportunity Map, to identify useful and actionable knowledge quickly and easily from the discovered rules. The framework is inspired by the House of Quality from Quality Function Deployment (QFD) in Quality Engineering. It associates discovered rules, related summarized data and data distributions with the application objective using an interactive matrix. Combined with drill down visualization, integrated visualization of data distribution bars and rules, visualization of trend behaviors, and comparative analysis, the Opportunity Map allows users to analyze rules and data at different levels of detail and quickly identify the actionable knowledge and opportunities. The proposed framework represents a systematic and flexible approach to rule analysis. Applications of the system to large-scale data sets from our industrial partner have yielded promising results.

Using gene expression programming to construct sentence ranking functions for text summarization

In this paper, we consider the automatic text summarization as a challenging task of machine learning. We proposed a novel summarization system architecture which employs Gene Expression Programming technique as its learning mechanism. The preliminary experimental results have shown that our prototype system outperforms the baseline systems.

An intelligent data mining system for drop test analysis of electronic products

Drop testing is one common method for systematically determining the reliability of portable electronic products under actual usage conditions. The process of drop testing, interpreting results, and implementing design improvements is knowledge-intensive and time-consuming, and requires a great many decisions and judgments on the part of the human designer. To decrease design cycles and, thereby, the time to market for new products, it is important to have a method for quickly and efficiently analyzing drop test results, predicting the effects of design changes, and determining the best design parameters. Recent advances in data mining have provided techniques for automatically discovering underlying knowledge from large amounts of experimental data. In this paper, an intelligent data mining system named decision tree expert (DTE) is presented and applied to drop testing analysis. The rule induction method in DTE is based on the C4.5 algorithm. In our preliminary experiments, concise and accurate conceptual design rules were successfully generated from drop test data after incorporation of domain knowledge from human experts. The data mining approach is a flexible one that can be applied to a number of complex design and manufacturing processes to reduce costs and improve productivity.

Optimization of a high-speed placement machine using tabu search algorithms

Combinatorial optimization represents a wide range of real-life manufacturing optimization problems. Due to the high computational complexity, and the usually high number of variables, the solution of these problems imposes considerable challenges.This paper presents a tabu search approach to a combinatorial optimization problem, in which the objective is to maximize the production throughput of a high-speed automated placement machine. Tabu search is a modern heuristic technique widely employed to cope with large search spaces, for which classical search methods would not provide satisfactory solutions in a reasonable amount of time. The developed TS strategies are tailored to address the different issues caused by the modular structure of the machine.

Design-to-Manufacturing Information Management for Electronics Assembly

This paper addresses design to manufacturing (DTM) for electronics assembly from several different perspectives. First, a working definition for DTM is proposed, and the distinction is made between DTM and the more commonly known design for manufacturability. Following an overview of surface mount technology assembly processes, DTM information management is introduced in terms of its data requirements and underlying decision and planning problems. Evaluation criteria are discussed, and specific requirements for a state-of-the-art DTM system are highlighted. The experiences of one Motorola factory are presented in a brief case study, which covers process mapping, system design and benchmarking, and installation and configuration activities. Benefits of an efficient DTM system are discussed in terms of the improvements in production time, engineering time, and product or process quality. Finally, a summary of future trends for DTM is given.

Simulation software for surface mount assembly

SMT Sim is a simulation program specially developed for surface mount assembly. Detailed models have been constructed for studying feeder setup and chip placement operations. This software tool can be used for estimating the production cycle time for a particular set of boards and for evaluating the impact of various production policies on a surface mount factorys throughput. Following some introductory comments regarding the need for developing a tool such as SMT Sim, this paper presents an overview of the functional elements of the software, i.e., the routines for managing input data, outputs, and process models. The object-based code libraries of the SMT Sim software are also discussed. Use of the software is explained by means of an example simulation run, from build plan to simulated production report. Two types of applications are addressed: benchmarking setup procedures and analyzing assembly cost.

Optimal versus heuristic scheduling of surface mount technology lines

This paper presents and compares an exact and a heuristic approach for scheduling of printed wiring board assembly in surface mount technology (SMT) lines. A typical SMT line consists of several assembly stations in series and/or in parallel, separated by finite intermediate buffers. The objective of the scheduling problem is to determine the detailed sequencing and timing of all assembly tasks for each individual board, so as to maximize the lines productivity, which is defined in terms of makespan for a mix of board types. The limited intermediate buffers between stations result in a scheduling problem with machine blocking, where a completed board may remain on a machine and block it until a downstream machine becomes available. In addition, limited machine availability due to scheduled downtimes is considered. The exact approach is based on a mixed integer programming formulation that can be used for optimization of assembly schedules by using commercially available software for integer programming, whereas the heuristic approach is designed as a combination of tabu search and a set of dispatching rules. Numerical examples modelled after real-world SMT lines and some computational results are provided to illustrate and compare the two approaches.