Homeworkยถ

Readingยถ

We recommend revisiting the papers by Petousis et al. on the dielectric constants datatset [1] and by Robertson on high-\(\kappa\) oxides [2]. Even if you read a paper once before, it never hurts to do another pass! We hope that by doing so, you get a better sense of which properties should be the outputs and which might be good inputs.

Programmingยถ

We recommend reviewing the ML exercises from today, which covered quite a lot of information in a very short amount of time.

If you want to try applying your knowledge, see if you can build a vanilla linear model for your dielectric materials. Be sure to consider:

  • Is this a regression or classification problem?

  • What are the inputs and what are the output(s)?

  • Where will I get the data? Hint: you should already have them!

  • What tools will help me? Hint: pandas and scikit-learn.

See the accompanying notebook for some guided exercises, if you would find that helpful.

Referencesยถ

1

Ioannis Petousis, David Mrdjenovich, Eric Ballouz, Miao Liu, Donald Winston, Wei Chen, Tanja Graf, Thomasย D. Schladt, Kristinย A. Persson, and Fritzย B. Prinz. High-throughput screening of inorganic compounds for the discovery of novel dielectric and optical materials. Scientific Data, 4(1):160134, 2017. doi:10.1038/sdata.2016.134.

2

J.ย Robertson. High dielectric constant oxides. The European Physical Journal Applied Physics, 28(3):265โ€“291, 2004. doi:10.1051/epjap:2004206.