A working knowledge of Artificial Neural Networks is rapidly becoming critical for navigating the modern world. Although the last few years have seen an explosion of the use of these tools in research, and there are many do-it-yourself articles on the web, they have not yet filtered down to wide implementation in introductory courses. I report here on my integration of machine learning activities into a general education course on galaxies and cosmology. I describe four lab activities for image classification, and I reflect on the strengths and weaknesses of using these tools in the context of online instruction during the 2020-21 pandemic academic year.

1.
There are some fantastic introductory videos and transcripts at https://www.3blue1brown.com/topics/neural-networks, accessed Aug. 9, 2021, that explain the basic concepts and methods of machine learning.
2.
The site medium.com has many articles, such as https://carterrhea93.medium.com/astronomical-images-for-machine-learning-applications-b50e7f298337, accessed May 18, 2021. Github and Kaggle also provide many, many resources. Any details I might explain here would probably be obsolete by the time this article is printed.
3.
http://ads.harvard.edu/, accessed June 1, 2021.
4.
J. P.
Davis
and
W. A.
Price
, “
Deep learning for teaching university physics to computers
,”
Am J. Phys.
85
,
311
(
April
2017
).
5.
https://keras.io/, accessed June 2, 2021.
6.
https://www.tensorflow.org/, accessed June 1, 2021.
8.
C. J.
Lintott
 et al, “
Galaxy Zoo: Morphologies derived from visual inspection of galaxies from the Sloan Digital Sky Survey
,”
Mon. Not. R. Astron. Soc.
389
,
1179
(
2008
).
9.
S.
Slater
,
T.
Slater
, and
D. J.
Lyons
,
Engaging in Astronomical Inquiry
(
W.H. Freeman & Company
,
2010
).
10.
E.
Prather
,
T.
Slater
,
J.
Adams
,
J.
Bailey
,
L.
Monowar-Jones
, and
J.
Dostal
, “
Research on a Lecture-Tutorial approach to teaching introductory astronomy for non-science majors
,”
Astron. Educ. Rev.
3
(
2
),
122
(
2004
)
11.
https://www.sdss.org/, accessed Aug. 9, 2021.
13.
A Jupyter Notebook is a web-browser-based interactive computing environment that uses the Python language to carry out programming tasks. It also supports HTML and LaTeX markup languages to present text and image-based commentary on and explanations of the material in the notebook. The result is an interactive web page with images, graphics, text, equations, and active Python code. Jupyter Notebooks are used extensively in programming, physics, and data science research and commerce.
AAPT members receive access to The Physics Teacher and the American Journal of Physics as a member benefit. To learn more about this member benefit and becoming an AAPT member, visit the Joining AAPT page.