Summary:  On  May  10,  2017  an  internal  symposium  titled  Machine  Learning  in Science and Engineering was held at Carnegie Mellon University to identify ways in which  these  computational  tools  are  advancing  a  diversity  of  fields.  In  a  very exciting  day  that  featured  23  research  talks  by  CMU  faculty  and  17  posters  were presented by graduate students and post-‐doctoral associates on topics ranging from the  search  for  dark  matter  in  the  universe  to  development  of  tools  for  product design.  Based  on  the  strong  response  at  CMU  with  over  120  registered  across computer  science,  engineering,  and  science,  the  organizers  are  planning  a  larger, open conference on June 6-‐8, 2018 at the CMU campus in Pittsburgh.  
This  conference  will  be  co-‐organized  with  Georgia  Tech.  While  both  CMU  and  GT will be strongly represented, the conference is expected to have 350 attendees from academia, government (NIST, Department of Energy, Department of Defense), and a diversity   of   companies   that   are   implementing   data   sciences   and   artificial intelligence in their operations.  
On  the  morning  of  June  6,  short  courses  in  a  variety  of  topics  will  be  offered  for attendees on topics ranging from  basics of  machine  learning to advanced topics  in computer  vision.  The  afternoon  of  June  6  will  feature  plenary  talks  from  leading researchers  across  disciplines  as  well  as  a  panel  discussion  on  industry  and government  interests  in  these  areas.  June  7-‐8  will  have  two  days  of  focused programming,  allowing  attendees  to  learn  about  the  state-‐of-‐the-‐art  in  machine learning applied across science and engineering fields.  
Intellectual Merit: While machine learning has revolutionized many areas of biological and biomedical research, its impact across the physical sciences and engineering is at an early stage. This conference will bring together researchers in a diversity of STEM areas focused on applying machine learning to problems of 
fundamental or applied nature. Presentations will focus on adapting existing 
machine  learning  methods  to  current  research  areas,  developing  new  machine learning   algorithms   specific   to   science   and   engineering,   and   identifying   new frontiers of research that may only be pursued using a data-‐centric approach. The conference will reflect the broad technical scope of the 2017 symposium while also offering attendees focused short courses taught by experts in machine learning on a variety of cutting-‐edge tools that are critical in advancing these fields.

Broader  Impacts:  Disseminating  machine  learning  methods  across  science  and engineering could have broad implications for US research. A conference grant from NSF  is  sought  to  support  general  conference  costs,  attendance  of  students  from regions outside Pittsburgh, as well as involving participation of students and faculty from  underrepresented  groups.  CMU  has  strong  relationships  with  a  number  of minority-‐serving  institutions,  such  as  Morgan  State  University  and  University  of Texas – El Paso.

The  organizers,  themselves  represent  a  very  diverse  group.   Beyond  intellectual diversity, the main organizers include 3 women.   The tracks are each co-‐led by a CMU and a GT faculty member, and among these there are 6 women, 2 of whom are also  URMs.    The  organizers  are  committed  to  promoting  participation  among underrepresented groups, junior researchers and students, and including tutorials to widen accessibility to as large a group of attendees, as possible.   We will make sure  the  symposium  is  advertised  broadly,  including  HBCs,  especially  in  the
Pittsburgh and Atlanta regions.