Data mining: A tool for detecting cyclical disturbances in supply networks

Abstract

Disturbances in supply chains may be either exogenous or endogenous. The ability automatically to detect, diagnose, and distinguish between the causes of disturbances is of prime importance to decision makers in order to avoid uncertainty. The spectral principal component analysis (SPCA) technique has been utilized to distinguish between real and rogue disturbances in a steel supply network. The data set used was collected from four different business units in the network and consists of 43 variables; each is described by 72 data points. The present paper will utilize the same data set to test an alternative approach to SPCA in detecting the disturbances. The new approach employs statistical data pre-processing, clustering, and classification learning techniques to analyse the supply network data. In particular, the incremental k-means clustering and the RULES-6 classification rule-learning algorithms, developed by the present authors' team, have been applied to identify important patterns in the data set. Results show that the proposed approach has the capability automatically to detect and characterize network-wide cyclical disturbances and generate hypotheses about their root cause.

Self-Adjusting Variable Neighborhood Search Algorithm for Near-Optimal k-Means Clustering

“…Comparative results for Mopsi-Joensuu data set. 6014 data vectors in R 2 , k = 30 clusters, time limitation 5 s. 3.43194 × 10 10 3.49405 × 10 10 3.44496 × 10 10 3.44140 × 10 10 1.64 360 × 10 8 GREEDY 3 3.43195 × 10 10 3.49411 × 10 10 3.44474 × 10 10 3.44140 × 10 10 1.71131 × 10 8 GREEDY 5 3.43195 × 10 10 3.48411 × 10 10 3.44663 × 10 10 3.44141 × 10 10 1.65153 × 10 8 GREEDY 7 3.42531 × 10 10 3.47610 × 10 10 3.44091 × 10 10 3.43504 × 10 10 1.76023 × 10 8 GREEDY 10 3.42560 × 10 10 3.48824 × 10 10 3.45106 × 10 10 3.43573 × 10 10 2.36526 × 10 8 GREEDY 12 3.42606 × 10 10 3.48822 × 10 10 3.44507 × 10 10 3.43901 × 10 10 1.68986 × 10 8 GREEDY 15 3.42931 × 10 10 3.47817 × 10 10 3.43874 × 10 10 3.43901 × 10 10 8.31510 × 10 7 GREEDY 20 3.42954 × 10 10 3.48826 × 10 10 3.44186 × 10 10 3.43905 × 10 10 1.28972 × 10 8 GREEDY 25 3.43877 × 10 10 3.44951 × 10 10 3.43982 × 10 10 3.43907 × 10 10 2.57320 × 10 7 GREEDY 30 3.43900 × 10 10 3.48967 × 10 10 3.45169 × 10 10 3.43979 × 10 10 1.93565 × 10 8 GH-VNS1…”

Assessing supplier capabilities to exploit building information modelling

“…This latter approach is purely based on an objective function and is not so prone to subjectivity regarding the existence of certain attributes as per the maturity definitions. While the K-means approach may be modified, for example, to allow for automatic selection of the optimum number of clusters (Afify et al , 2007) or alternatives clustering approaches selected (Brusco et al , 2012) in its classic form it still has considerable utility due to its ease of use and simplicity (Ratrout, 2011) and is particularly pertinent to data sets such as ours, where which we wish to partition through establishing virtual cluster centers (Brusco et al , 2012).…”

‘Bullwhip’ and ‘backlash’ in supply pipelines

“…This technique decomposes a time series into sinusoidal components at different frequencies with each frequency 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 Fourier analysis or forms of it (Thornhill and Naim, 2006;Afify et al, 2007) has previously been used to detect 'bullwhip' and hence has potential for 'backlash' detection. 'Backlash' and 'Bullwhip' are related as given in Figure 5 and therefore if we could profile them together, this could be used as a characteristic signature for detecting the 'backlash' effect.…”