The Effect of Machine Learning Techniques on Materials Informatics

[tonyfast]

The problem

A common question that has been arising

What situations can we used supervised learning?

Supervised and Unsupervised Learning

• Principal Component Analysis is a common unsupervised learning technique.
• Linear Discriminant Analysis is a simple supervised learning technique.

  Relevance of the problem

Experimental materials science data is generated under a set of processing and sensing conditions. There is an intrinsic uncertainty in processing conditions. However, we can assume the sensing conditions are the same for a realization of a structure.

Hypothesis

Independent realizations of structures can be classified as like categories because the processing and sensing conditions are the near uniform for ONE realization.

[tonyfast]

Here is a post that moves closer to this question.