Stefan Simeonov Dimov

Selection of K in K-means clustering

Abstract The K-means algorithm is a popular data-clustering algorithm. However, one of its drawbacks is the requirement for the number of clusters, K, to be specified before the algorithm is applied. This paper first reviews existing methods for selecting the number of clusters for the algorithm. Factors that affect this selection are then discussed and a new measure to assist the selection is proposed. The paper concludes with an analysis of the results of using the proposed measure to determine the number of clusters for the K-means algorithm for different data sets.

Selective laser sintering: Applications and technological capabilities:

Abstract Rapid prototyping (RP) models are no longer used only for design verification. Currently, parts built utilizing layer manufacturing technology can be employed as functional prototypes and as patterns or tools for different manufacturing processes such as vacuum casting, investment casting, injection moulding, die casting and sand casting. The selective laser sintering (SLS) process is becoming widely used in industry, its main advantage over other RP processes being that it can produce parts in a wide range of materials. This paper describes applications of some of the materials available with the SLS process and, by means of case studies, explains the technological capabilities of this process.

An incremental K-means algorithm

Abstract Data clustering is an important data exploration technique with many applications in engineering, including parts family formation in group technology and segmentation in image processing. One of the most popular data clustering methods is K-means clustering because of its simplicity and computational efficiency. The main problem with this clustering method is its tendency to coverge at a local minimum. In this paper, the cause of this problem is explained and an existing solution involving a cluster centre jumping operation is examined. The jumping technique alleviates the problem with local minima by enabling cluster centres to move in such a radical way as to reduce the overall cluster distortion. However, the method is very sensitive to errors in estimating distortion. A clustering scheme that is also based on distortion reduction through cluster centre movement but is not so sensitive to inaccuracies in distortion estimation is proposed in this paper. The scheme, which is an incremental version of the K-means algorithm, involves adding cluster centres one by one as clusters are being formed. The paper presents test results to demonstrate the efficacy of the proposed algorithm.

Rapid prototyping and rapid tooling: the key enablers for rapid manufacturing

Abstract Rapid manufacturing is a new mode of operation that can greatly improve the competitive position of companies adopting it. The key enabling technologies of rapid manufacturing are rapid prototyping (RP) and rapid tooling (RT). This paper classifies the existing RP processes and briefly describes those with actual or potential commercial impact. The paper then discusses five important RP applications: building functional prototypes, producing casting patterns, making medical and surgical models, creating artworks and fabricating models to assist engineering analysis. Finally, the paper gives an overview of indirect and direct RT methods for quickly producing up to several thousand parts together with examples illustrating different applications of RT.

An efficient algorithm for automatic knowledge acquisition

This paper presents a new algorithm for extracting IF-THEN rules from examples. The algorithm employs an efficient rule searching method and a simple metric for assessing rule generality and accuracy. The paper illustrates step by step the operation of the algorithm and discusses its performance on the IRIS data classification problem.

An Algorithm for Incremental Inductive Learning

Abstract This paper describes RULES-4, a new algorithm for incremental inductive learning from the ‘RULES’ family of automatic rule extraction systems. This algorithm is the first incremental learning system in the family. It has a number of advantages over well-known non-incremental schemes. It allows the stored knowledge to be updated and refined rapidly when new examples are available. The induction of rules for a process planning expert system is used to illustrate the operation of RULES-4 and a bench-mark pattern classification problem employed to test the algorithm. The results obtained have shown that the accuracy of the extracted rule sets is commensurate with the accuracy of the rule set obtained using a non-incremental algorithm.

Micromilling: material microstructure effects

Abstract Micromilling is one of the technologies that is currently widely used for the production of microcomponents and tooling inserts. To improve the quality and surface finish of machined microstructures the factors affecting the process dynamic stability should be studied systematically. This paper investigates the machining response of a metallurgically and mechanically modified material. The results of micromilling workpieces of an Al 5000 series alloy with different grain microstructures are reported. In particular, the machining response of three Al 5083 workpieces whose microstructure was modified through a severe plastic deformation was studied when milling thin features in microcomponents. The effects of the material microstructure on the resulting part quality and surface integrity are discussed and conclusions made about its importance in micromilling. The investigation has shown that through a refinement of material microstructure it is possible to improve significantly the surface integrity of the microcomponents and tooling cavities produced by micromilling.

Rapid tooling applications of the selective laser sintering process

The selective laser sintering (SLS) process is one of the leading rapid prototyping techniques. This paper presents two rapid tooling (RT) methods based on the SLS process. The first method employs the SLS process to build tooling inserts in copper polyamide that can be used for fabrication of a limited number of pre‐production parts in the same material and manufacturing process as the final production parts. The second method, the RapidToolTM process, is a RT solution for manufacture of pre‐production and production tools for injection moulding and die‐casting. The paper also discusses the applications and limitations of these RT methods.

Micromilling strategies: optimization issues

Abstract Micromilling is one of the technologies widely used to manufacture microstructures and tooling inserts for microinjection moulding and hot embossing. A number of manufacturing constraints remain that limit the application of this technology. One of these constraints is that the existing machining strategies are not appropriate for the manufacture of features that are common in micro parts. This paper discusses an approach for optimizing these strategies. The aim is to provide users of computer aided manufacturing (CAM) systems with tools enabling them to generate cutter paths that take into account the specific conditions arising during micromilling. The paper studies the advantages and disadvantages of using different machining strategies for micromilling and then verifies their capabilities experimentally. Also, an approach is proposed for storing and re-using expert knowledge about micromachining strategies associated with different feature types.

RULES-5: a rule induction algorithm for classification problems involving continuous attributes

Abstract This paper presents RULES-5, a new induction algorithm for effectively handling problems involving continuous attributes. RULES-5 is a ‘covering’ algorithm that extracts IF-THEN rules from examples presented to it. The paper first reviews existing methods of rule extraction and dealing with continuous attributes. It then describes the techniques adopted for RULES-5 and gives a step-by-step example to illustrate their operation. The paper finally gives the results of applying RULES-5 and other algorithms to benchmark problems. These clearly show that RULES-5 generates rule sets that are more accurate than those produced by its immediate predecessor RULES-3 Plus and by a well-known commercially available divide-and-conquer machine learning algorithm.