On the last Machine Learning session class, I tried to explain about how to get one of object features from handwriting alphabetic image. This step called feature extraction in machine learning.
Getting the valid and accurate feature is very important thing to build strong data train on machine learning. Because all of data train will imply to output system performance. Always remember that all methods of machine learning are depend in training process based on data train which we got from object feature extraction step.
One of simple digital image feature is region of interest area. You can get this feature just by count whole of pixels which shape the alphabetic character on image. [continued …]