AIMS South Africa Newsletter November 2024

29 views
Skip to first unread message

Linda Camara

unread,
Dec 12, 2024, 9:28:24 AM12/12/24
to Mailing List: News and Announcments
  

AIMS Celebrates Emerging Scientists at the Recognition of Achievement Ceremony

  

AIMS South Africa was delighted to celebrate the graduation of its January 2024 intake from the Master of Science in Mathematical Sciences programme. The Recognition of Achievement Ceremony, held on 28 November 2024, honoured the hard work and outstanding achievements of these dedicated students.

  

This latest cohort consists of 29 students, including 7 women, hailing from 9 African countries, with 11 students from South Africa. This brings the total number of AIMS South Africa graduates to 1,089, of whom 35% are women. The students on the AI for Science Stream, totaling 11 in this cohort, were funded by Google DeepMind, underscoring the programme’s commitment to developing cutting-edge AI talent in Africa.

  
  
  
  
  
Read more
  

Training Africa’s ‘computer conservationists’

  

We’re excited to share how Dr Emmanuel Dufourq, AIMS-Carnegie Junior Research Chair at AIMS South Africa, is leading the way in training Africa’s future 'computer conservationists.' Featured in Nature, this inspiring piece highlights his work in developing accessible AI solutions for wildlife research in under-resourced contexts. Dive into the full story to see how AI is making an impact in conservation efforts across Africa. 

  
  
  
Read more

Folding Numbers: Mathematics meets Origami

  
  
  

On 14 November AIMS South Africa hosted a public lecture titled “Folding Numbers – Mathematics Meets Origami,” presented by Prof. Michael Winckler from Heidelberg University. This two-part event – lecture and hands-on workshop – showcased intricate connections between the ancient Japanese art of paper folding and modern mathematics.  

  

It attracted more than 100 people, including AIMS students, staff, visiting lecturers, industry professionals, artists, a delegation from the German Embassy in Pretoria, and retirees who value lifelong learning from Muizenberg and the surrounding communities.

  
Read more
  
  

HeiAIMS

  
  
  
  

From November 11 to 15, Prof. Michael Winkler and Sarah Steinbach from the Interdisciplinary Center for Scientific Computing (IWR) at Heidelberg University visited AIMS for the HeiAIMS midterm review meeting. The purpose of the review was to reflect on lessons learned, make plans for the future of the programme, and discuss proposals. HeiAIMS is a three-year project aimed at strengthening and developing the cooperation between Heidelberg University and AIMS.

  

Student Updates

  

Structured Master’s students completed: Difference and Differential Equations in Mathematical Modelling (Jacek Banasiak, Pretoria) and SAGE (Dr Evans Doe Ocansey, RISC Software GmbH).

  

AI for Science Stream students completed: Mathematical Problem Solving (Jacques Rabie, Cape Town) and Reinforcement Learning (Arnu Pretorius, InstaDeep)

  
  

Machine Learning for Ecology Group Updates

Dr Lorene Jeantet and Dr Emmanuel Dufourq, in collaboration with Damien Chevallier from CNRS France, published a new article in Journal of Experimental Biology entitled “Automatic identification of the endangered Hawksbill sea turtle behaviour using deep learning and cross-species transfer learning”.   

  

In this paper, we show that data from another species (e.g., green turtle or human) can be used for pre-training a model to identify behaviours from accelerometer data. The pre-trained model can then be fine-tuned for the target species (e.g., hawksbill turtle), achieving better results than training solely on the target species.  

  

This study was part of Kukhanya Zondo's research project for his Master's in Mathematical Sciences at AIMS South Africa, where he graduated with distinction in November 2023! He is now one of the co-authors of this article.

  
  
  

Kukhanya celebrates with group members Matthew and Milanto.

  

Deep Learning Tackles Animal ID

PhD student Emmanuel Kabuga led a publication that used deep learning to automatically identify whether a pair of images was of the same animal or not. Deep learning is widely used for species identification but discriminating between individual animals of the same species is a much harder problem. Emmanuel tested his approach on four species: humpback whales, bottlenose dolphins, harbour seals, and the local western leopard toad, a species found only in the Western Cape (and not far from the AIMS South Africa campus). The paper, which appeared in Ecosphere, was co-authored with AIMS associate research fellows 

  
  

Bubacarr Bah and Ian Durbach, and was featured in the Ecological Society of America's monthly roundup of five research articles published across its esteemed journals.

  
Read more
  

Visit to Sol Plaatje University

On 18 November, as part of the Sol Plaatje University’s Faculty of Science’s Research Colloquium, Dr Karin-Therese Howell, AIMS Academic Director, had the opportunity to share her research, talk about what we do at AIMS South Africa and engage with honours students presenting their work.   

  

The colloquium marked a nostalgic full circle for Prof. Howell, who began her career at the National Institute for Higher Education, which later evolved into Sol Plaatje University. She expressed her gratitude to Dr Meenakshi Devi, Dr Dimpho Mothibi, and Prof. Nhamo Chaukura for their warm hospitality, and to Prof. Martin Ntwaeaborwa, the Dean of Sciences, for the invitation to participate.

  
  
  

Advancing science communication in Africa:
Reframing Afrocentric perspectives on inclusion, capacity-building
and language plurality

From 18 to 20 November, Dr Rejoyce Gavhi-Molefe, AIMS House of Science Manager, attended and presented an insightful talk at the Communicating Discovery Science Symposium, hosted by the Centre for Research on Evaluation, Science and Technology (CREST) at Stellenbosch University. The three-day symposium was attended by researchers and practitioners in science communication. They shared various ways to improve public engagement with Discovery Science, also known as Basic Science.

  
  
  
Read more
  

AIMS Women in STEM:
Inspiring Insights from Dr Zara Randramanakoto

  
  
  
Read more

On 22 November 2024, the House of Science’s Women in STEM mentoring programme featured an inspiring session with Dr Zara Randramanakoto, an Astronomer and co-founder of Ikala STEM (Women in STEM – Madagascar). During her engaging and interactive session, which female AIMS students attended, Dr Randramanakoto shared her remarkable journey from Madagascar to become a leading researcher at the South African Astronomical Observatory. Her journey included personal and professional experiences that provided valuable life lessons.

  

Exploring Shweshwe Patterns

  
  
  

In a workshop, held on 16 November 2024, with 40 learners from Intlanganiso Secondary School in Khayelitsha, we embarked on a collaborative exploration of the idea of algorithms. Learners, working in groups alongside AIMS Master's students, deconstructed and reconstructed traditional South African Shweshwe fabric patterns. Together, they identified basic shapes, patterns of repetition, and symmetries used in the Shweshwe designs, and collectively devised an “algorithm” to build the patterns on the fabric. This shared experience is intended to inspire new ideas for future designs and help learners structure their approach to creating a digital design.

  
  
Read more

Experimenting with reverse-teaching: Learners guide teachers

  
  
  

On 9 November, AIMS South Africa held its first-ever Computational Thinking course for teachers, led by an expert facilitator from SAP and AIMSSEC staff. This event was groundbreaking, as learners from Masibambane and Lavender Hill High Schools — who recently completed algorithmic thinking courses at AIMSSEC — supported and even taught alongside the facilitators, introducing teachers to a “learners-guide-teachers” model rarely seen in South Africa’s education system. Throughout the day, teachers explored the essential skills of algorithmic construction, deconstruction, and reconstruction, learning to break down complex problems into manageable steps and analyse existing solutions.

  
  
Read more
  

Teaching tip for lecturers!

  
  

Strategies for group work in education often fall into routine patterns, like always pairing students or grouping them without a clear pedagogical strategy. Here are two engaging methods to enhance group interactions during lectures:  

  
  

Listening Triads: Assign students into groups of three with specific roles: a talker who shares thoughts, a questioner who probes deeper, and a recorder who summarizes the discussion. Rotate roles to vary participation throughout the module.  

  
  

Envoys: Students first work in their initial groups to complete a task. Subsequently, one member from each group, the envoy, joins another group to exchange findings and discuss similarities or differences. Envoys then return to their original groups to relay insights.

  
  

 

 

ANNOUNCEMENTS

We are hiring!

  
  

We have open positions for:

  

A Machine Learning Researcher: Earth Observation & Ecology (open until filled.) 

  

AIMS IT Support Manager (Application Deadline:  31 December 2024)  

  

AIMS Finance and Technology Manager (Application Deadline:
31 December 2024)  

  
  

AIMS IT Manager (Application Deadline: 31 December 2024)  

  

Please see our applications page for more details!

  
Read more
  
  

Click here to unsubscribe from our mailing list.

  
Reply all
Reply to author
Forward
0 new messages