Accelerating skills development in Data science and AI at scale
At the Monash Data Science and AI platform, we believe that upskilling our research community and building a workforce with data science skills are key to accelerating the application of data science in research. To achieve this, we create and leverage new and existing training capabilities...
Keywords: AI, machine learning, eresearch skills, training, train the trainer, volunteer instructors, training partnerships, training material
Accelerating skills development in Data science and AI at scale
https://zenodo.org/records/4287746
https://dresa.org.au/materials/accelerating-skills-development-in-data-science-and-ai-at-scale-2d8a65fa-f96e-44ad-a026-cfae3f38d128
At the Monash Data Science and AI platform, we believe that upskilling our research community and building a workforce with data science skills are key to accelerating the application of data science in research. To achieve this, we create and leverage new and existing training capabilities within and outside Monash University. In this talk, we will discuss the principles and purpose of establishing collaborative models to accelerate skills development at scale. We will talk about our approach to identifying gaps in the existing skills and training available in data science, key areas of interest as identified by the research community and various sources of training available in the marketplace. We will provide insights into the collaborations we currently have and intend to develop in the future within the university and also nationally.
The talk will also cover our approach as outlined below
• Combined survey of gaps in skills and trainings for Data science and AI
• Provide seats to partners
• Share associate instructors/helpers/volunteers
• Develop combined training materials
• Publish a repository of open source trainings
• Train the trainer activities
• Establish a network of volunteers to deliver trainings at their local regions
Industry plays a significant role in making some invaluable training available to the research community either through self learning platforms like AWS Machine Learning University or Instructor led courses like NVIDIA Deep Learning Institute. We will discuss how we leverage our partnerships with Industry to bring these trainings to our research community.
Finally, we will discuss how we map our training to the ARDC skills roadmap and how the ARDC platforms project “Environments to accelerate Machine Learning based Discovery” has enabled collaboration between Monash University and University of Queensland to develop and deliver training together.
contact@ardc.edu.au
Tang, Titus
AI, machine learning, eresearch skills, training, train the trainer, volunteer instructors, training partnerships, training material
Data Fluency: a community of practice supporting a digitally skilled workforce
This presentation showcases the impact of the Monash Data Fluency Community of Practice upon digitally skilled Graduate Research students involved as learners and instructors in the program. The strong focus on building community to complement training, has fostered an environment of learning,...
Keywords: skills, training, eresearch skills, data skills, online learning, pedagogy, train the trainer, digitally skilled workforce, training material
Data Fluency: a community of practice supporting a digitally skilled workforce
https://zenodo.org/records/4287752
https://dresa.org.au/materials/data-fluency-a-community-of-practice-supporting-a-digitally-skilled-workforce-b911a1a8-0331-496e-95a6-0015a12acc34
This presentation showcases the impact of the Monash Data Fluency Community of Practice upon digitally skilled Graduate Research students involved as learners and instructors in the program. The strong focus on building community to complement training, has fostered an environment of learning, networking and sharing of expertise. Hear what the Graduate research students have to say about the value of skills training and how it has impacted their research; how the community has enabled them to network with a broad range of researchers and affiliate partner groups they would not ordinarily be in contact with; how their research journey has been enhanced by working as part of a multi-disciplinary team, as well as sharpening their teaching skills.
The rapid refocus from face - face to online delivery, as a result of the pandemic, highlights the importance of the multi-faceted online approach including workshops, drop-in sessions, SLACK chat and online learning resources. As a result of the shift to online, the range of strategic external partner/affiliate groups has extended and demand for workshops and drop-ins has increased. Learn how the instructors have altered their pedagogical approach to engage workshop and drop-in participants; how they have overcome some of the challenges of facilitating in an online environment; and how this is preparing them to become part of a digitally skilled workforce.
contact@ardc.edu.au
Groenewegen, David (orcid: 0000-0003-2523-1676)
skills, training, eresearch skills, data skills, online learning, pedagogy, train the trainer, digitally skilled workforce, training material
CheckEM User Guide
CheckEM is an open-source web based application which provides quality control assessments on metadata and image annotations of fish stereo-imagery. It is available at marine-ecology.shinyapps.io/CheckEM. The application can assess a range of sampling methods and annotation data formats for...
Keywords: stereo-video, fish, annotation
CheckEM User Guide
https://globalarchivemanual.github.io/CheckEM/articles/manuals/CheckEM_user_guide.html
https://dresa.org.au/materials/checkem-user-guide
CheckEM is an open-source web based application which provides quality control assessments on metadata and image annotations of fish stereo-imagery. It is available at marine-ecology.shinyapps.io/CheckEM. The application can assess a range of sampling methods and annotation data formats for common inaccuracies made whilst annotating stereo imagery. CheckEM creates interactive plots and tables in a graphical interface, and provides summarised data and a report of potential errors to download.
brooke.gibbons@uwa.edu.au
Brooke Gibbons
stereo-video, fish, annotation
EventMeasure Annotation Guide
EventMeasure annotation guide for baited remote underwater stereo video systems (stereo-BRUVs) for count and length
Keywords: fish, stereo-video, annotation
EventMeasure Annotation Guide
https://globalarchivemanual.github.io/CheckEM/articles/manuals/EventMeasure_annotation_guide.html
https://dresa.org.au/materials/eventmeasure-annotation-guide
EventMeasure annotation guide for baited remote underwater stereo video systems (stereo-BRUVs) for count and length
tim.langlois@uwa.edu.au
Brooke Gibbons
Tim Langlois
Claude Spencer
fish, stereo-video, annotation
Stereo-video workflows for fish and benthic ecologists
Stereo imagery is widely used by research institutions and management bodies around the world as a cost-effective and non-destructive method to research and monitor fish and habitats (Whitmarsh, Fairweather and Huveneers, 2017). Stereo-video can provide accurate and precise size and range...
Keywords: stereo-video, fish, sharks, habitats
Resource type: tutorial
Stereo-video workflows for fish and benthic ecologists
https://globalarchivemanual.github.io/CheckEM/index.html
https://dresa.org.au/materials/stereo-video-workflows-for-fish-and-benthic-ecologists
Stereo imagery is widely used by research institutions and management bodies around the world as a cost-effective and non-destructive method to research and monitor fish and habitats (Whitmarsh, Fairweather and Huveneers, 2017). Stereo-video can provide accurate and precise size and range measurements and can be used to study spatial and temporal patterns in fish assemblages (McLean et al., 2016), habitat composition and complexity (Collins et al., 2017), behaviour (Goetze et al., 2017), responses to anthropogenic pressures (Bosch et al., 2022) and the recovery and growth of benthic fauna (Langlois et al. 2020). It is important that users of stereo-video collect, annotate, quality control and store their data in a consistent manner, to ensure data produced is of the highest quality possible and to enable large scale collaborations. Here we collate existing best practices and propose new tools to equip ecologists to ensure that all aspects of the stereo-video workflow are performed in a consistent way.
tim.langlois@uwa.edu.au
Tim Langlois
Brooke Gibbons
Claude Spencer
stereo-video, fish, sharks, habitats