Implementation of Visual Clustering Strategy in Self-Organizing Map for Wear Studies Samples Printed Using FDM

In general, visual clusters are preferred over large data sets; this is an attempt to take advantage of cluster techniques reduce the mathematical complexity small sets. To identify possibility implementing clustering technique in a dataset, wear observations PLA/Cu composite samples printed using Fused Deposition Model (FDM) taken into consideration. study, Self Organizing Map (SOM) tool as non-supervised Neural Network (NN) used visualize data. Here, SOM combinations with vector quantification and projection or rank machinability parameters on new filament under different FDM conditions. The competitive layer will classify given machine (vectors) at any number dimensions may be several groups neurons. limitation map size which cannot exceed 1000 units training. However, for set consideration, extent these limits not affect performance. algorithm developed study provides outlet within acceptable range. addition, linear regression analysis carried out output response measure characteristics machining observation.