SOM-maps
Please refer the pdf made during the internship in france the motto of the internship is to test various techniques for outlier detection the idea still is extension of what I used the project of turbine efficiency predictions but this time, we deal with 3 other approaches as well 1- regression based approach 2- classification based approach 3- artificial neural networks based approach 4- unsupervised (Self organizing maps visual approach)
Note - We used soms as a visual matric to identify, hence all the results are based visually by looking at the points away or inside a map using dimensionality reduction techniques
application of SOM algorithms on Matlab/octave
run main.m to check and test the implementation