DISC: Data-Intensive Similarity Measure for Categorical Data

Desai, Aditya; Singh, Himanshu; Pudi, Vikram

doi:10.1007/978-3-642-20847-8_39

Cited by 16 publications

(5 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To assess the applicability of similarity measures for a specific CBR system the following evaluation criteria are used (derived from Boriah et al, 2008;Desai et al, 2011;Sulc and Rezankov a, 2014)…”

Section: Case Retrievalmentioning

confidence: 99%

Case-based reasoning for complexity management in Industry 4.0

Schott

Lederer

Eigner

et al. 2020

JMTM

View full text Add to dashboard Cite

PurposeIncreasingly, dynamic market environments lead to growing complexity in manufacturing and pose a severe threat for the competitiveness of manufacturing companies. Systematic guidance to manage this complexity, especially in the context of Industry 4.0 and the therewith rising trends such as digitalization and data-driven production optimization, is lacking. To address this deficit a case-based reasoning (CBR) system for providing knowledge about managing complexity in Industry 4.0 is presented.Design/methodology/approachFirst, the explicit knowledge representation for managing complexity in IT-based manufacturing is introduced. Second, the CBR process step to retrieve knowledge from an artificially composed case base with in total 70 cases of data-based complexity management in the context of Industry 4.0 is set out. Third, knowledge transfer alongside several maturity levels of information technology capabilities of manufacturing systems for reuse in new problem scenarios is introduced.FindingsThe paper comprises the conceptual approach for designing a CBR system to support data-based complexity management in manufacturing systems. Furthermore, the appropriateness of the CBR system to provide applicable knowledge for reducing and managing complexity in corporate practice is shown.Research limitations/implicationsThe presented research results are evaluated in the course of an embedded single case study and may therefore lack generalizability. Future research to test and enhance the appropriateness of the developed CBR system will strengthen the research contribution.Originality/valueThe paper provides a novel approach to systematically support knowledge transfer for data-based complexity management by transferring the well-known and established methodology of CBR to the rising application domain of manufacturing systems in the context of Industry 4.0.

show abstract

Section: Case Retrievalmentioning

confidence: 99%

Case-based reasoning for complexity management in Industry 4.0

Schott

Lederer

Eigner

et al. 2020

JMTM

View full text Add to dashboard Cite

show abstract

“…Conversely, authors in [19] make a comparative study of categorical variable encodings for predicting vehicle properties using neural networks. A similarity measure to handle categorical variables is introduced in [20] and validated in KNN regression problems over twelve datasets. In [21] authors propose a hybrid decision tree algorithm for mixed categorical and numerical regression analysis.…”

Section: Related Workmentioning

confidence: 99%

A Comparison Framework of Machine Learning Algorithms for Mixed-Type Variables Datasets: A Case Study on Tire-Performances Prediction

2020

View full text Add to dashboard Cite

Many engineering applications in the automotive, aeronautic, rubber, mechanics, and manufacturing industries collect multiple datasets measuring physical relations between input variables and performances for modeling purposes. The challenge relies on that such data is often highly dimensional, non-linear and contain mixed variables, i.e., numerical and categorical features, requiring specific algorithms and encoding schemes to perform regression task efficiently. Moreover, defining an appropriated similarity criterion for mixed-type data is a non-trivial task, especially when it is meant to be used in regression problems. This paper discusses the use of different machine learning algorithms for regression problems, involving mixed-type variables across multiple datasets. We use tire-related datasets as a case study to perform a rigorous, statistically founded comparison of different machine learning algorithms with encoding schemes to handle mixed variables in the prediction of tire-performances across multiple tirerelated datasets. Friedman's statistic and Nemenyi post-hoc tests are used to test the significance of performance differences between techniques and encoding strategies. Our contributions come as a series of recommendations for handling efficiently mixed-type variables while achieving high performances on regression tasks over multiple datasets. Furthermore, we provide a flexible and efficient similarity function between tires useful for tire comparison, prediction, and retrieval tasks. INDEX TERMS mixed-type variables, categorical encoding, Friedman Nemenyi, regression algorithms I. INTRODUCTION M ACHINE learning (ML) in engineering applications has grown in popularity during the last decades [1]-[5]. Many industrial applications use ML tools to build regression models for product design, performance optimization, variable design, fault detection, quality assessment, and others. For instance, the rubber industry, [6] uses non-linear least squares to estimate the tire-road friction coefficient for tire design. In automotive design, [7] employs support vector regression in structural optimization to vehicle crashworthiness design. More recently [8] performed thermodynamics compressor performance modeling for engine design with neural networks and non-linear support vector regression. A common characteristic of engineering data is its tabularlike structure, where rows represent data examples, which are themselves described as a mixture of numerical and categorical, i.e., mixed-type variables. For instance, in car crashworthiness design [9], vehicle structures must be designed to absorb crash energy through structural deformation as much as possible and attenuate the impact force to lower levels when impact occurs. The design variables are thickness related (measured in mm) and steel hardness types. For instance, the B-pillar inner, reinforce, the floor side and door beltline, are all numerical variables. However material design variables associated to the steel hardness, i.e., meal, medium, or high str...

show abstract

“…If data set contains only categorical attributes, algorithms like ROCK, K-modes [8], DISC [4] etc are available. Most of existing algorithms for mixed data types, have different distance measures for both numerical and categorical variables and combine those during clustering.…”

Section: Related Workmentioning

confidence: 99%

Mutual information based weighted clustering for mixed attributes

Jayabal

Ramanathan

2015

Proceedings of the Second ACM IKDD Conference on Data Sciences

View full text Add to dashboard Cite

There exists large number of clustering algorithms either for numeric or for categorical data sets. There are relatively less algorithms for clustering mixed attributes. This paper proposes Mutual Information based Weighted Clustering for Mixed Attributes (MI-WCMA) based on euclidean distance for numeric attributes, distance measure based on similarity for categorical attributes using rough sets and weights for features based on average mutual information. The metrics accuracy, silhouette width and kappa co-efficient are used for evaluation and comparison with existing algorithms.

show abstract

DISC: Data-Intensive Similarity Measure for Categorical Data

Cited by 16 publications

References 16 publications

Case-based reasoning for complexity management in Industry 4.0

Case-based reasoning for complexity management in Industry 4.0

A Comparison Framework of Machine Learning Algorithms for Mixed-Type Variables Datasets: A Case Study on Tire-Performances Prediction

Mutual information based weighted clustering for mixed attributes

Contact Info

Product

Resources

About