Welcome to the first issue of APL Machine Learning! Over the last few years, we have witnessed how machine learning and, more broadly, artificial intelligence (AI) are changing the world and impacting our everyday lives. Today, it is especially exciting that machine learning tools are utilized in scientific discovery, quite often of large complex problems that have been unsolved for many years. This is exemplified by recent successes of AI programs, such as AlphaFold, which can accurately predict 3D models of protein folding and structures that were an open problem in biology for 50 years.1 Of course, there are many other examples in different scientific disciplines, from physics, chemistry, materials science, and various branches of engineering, pharmacy, and medicine. The theme of applying machine learning for scientific discovery has found its way across most journals within the AIP Publishing portfolio, not only through individual papers but also through dedicated collections, special issues, and roadmaps. The time is ripe for launching a new open-access, specialized journal to serve the community, and I have the great privilege of being the first Editor-in-Chief of APL Machine Learning. I am equally privileged to work with an outstanding, diverse team of Associate Editors and Editorial Advisory Board members recognized for their contributions to the field.

Although APL in the name “APL Machine Learning” is not an acronym for applied physics, most of you will make this connection, and indeed it is part of the journal's vision. There are two natural ways to connect machine learning (or, more broadly, artificial intelligence) and applied physical sciences. I have already mentioned the first, applying machine learning techniques to accelerate scientific discovery in disciplines of applied physics. The second one is to use domain knowledge from different disciplines (e.g., materials science, engineering, biology, or neuroscience) to develop more efficient and functional machine learning/artificial intelligence systems. The journal’s scope covers both. We call the former “machine learning for applied physics” (ML for AP) and the latter “applied physics for machine learning” (AP for ML). This is represented in Fig. 1. The two perspective articles in this issue cover topics from these two research themes. The first perspective2 discusses the applicability of deep language models for materials science, with an overview of future challenges and opportunities. This is an exciting and important area of research as researchers seek to better understand the most suitable neural network architecture for specific domain applications. The second perspective3 provides a comprehensive overview of emerging memory devices, especially relevant for energy-efficient machine learning and novel computing systems, such as in-memory and neuromorphic computing. The development of such technologies is essential to meet high computing power demands for current and future AI applications.4 

FIG. 1.

Journal's scope consists of two elements: machine learning for applied physics (top) and applied physics for machine learning (bottom).

FIG. 1.

Journal's scope consists of two elements: machine learning for applied physics (top) and applied physics for machine learning (bottom).

Close modal

Notwithstanding the huge impact of the field, we fully recognize that the area is fast developing. Having the community in mind, we have devised data and code accessibility policy with the aim of actively supporting sustainable models of access to research that ensure the permanence, discoverability, and reuse of published work. All data, methods, and models should be well documented and described either in the main text of the article or supplementary material to provide the research community with enough transparency and detail to replicate the reported results. Therefore, APL Machine Learning requires that authors make any data, code, and additional supplementary material publicly available on a repository of the author’s choosing.

Finally, we know that needs of emerging research communities are continuously changing. We are committed to providing the optimal platform for disseminating exciting research and are eager to listen to the readers' and authors’ feedback on implementing new mechanisms to ensure we serve the community in the best possible way. Please feel free to reach out to us on our social media channels (our Twitter handle is @aplmachinelearn) or by email ([email protected]). We believe APL Machine Learning will be the leading venue for publishing the best and most exciting work, and we look forward to receiving your manuscripts!

1.
J.
Jumper
,
R.
Evans
,
A.
Pritzel
,
T.
Green
,
M.
Figurnov
,
O.
Ronneberger
,
K.
Tunyasuvunakool
,
R.
Bates
,
A.
Žídek
,
A.
Potapenko
,
A.
Bridgland
,
C.
Meyer
,
S. A. A.
Kohl
,
A. J.
Ballard
,
A.
Cowie
,
B.
Romera-Paredes
,
S.
Nikolov
,
R.
Jain
,
J.
Adler
,
T.
Back
,
S.
Petersen
,
D.
Reiman
,
E.
Clancy
,
M.
Zielinski
,
M.
Steinegger
,
M.
Pacholska
,
T.
Berghammer
,
S.
Bodenstein
,
D.
Silver
,
O.
Vinyals
,
A. W.
Senior
,
K.
Kavukcuoglu
,
P.
Kohli
, and
D.
Hassabis
,
Nature
596
,
583
(
2021
).
2.
Y.
Hu
and
M. J.
Buehler
,
APL Mach. Learn.
1
,
010901
(
2023
).
3.
P.
Manocci
,
M.
Farronato
,
N.
Lepri
,
L.
Cattaneo
,
A.
Glukhov
,
Z.
Sun
, and
D.
Ielmini
,
APL Mach. Learn.
1
,
010902
(
2023
).
4.
A.
Mehonic
and
A. J.
Kenyon
,
Nature
604
,
255
(
2022
).