Rainforest

Amazon rainforest: With 1,400 trees being felled every minute, 40,000 species of plant and the lives of more than 2,000 animal and bird species are at risk

Two MSc Business Analytics scholars joined the Data Science for Social Good fellowship (DSSGx) held at the University of Warwick recently to turn their skills to a good cause.  

Led by Professor of Operational Research & Systems, Juergen Branke, this was the fifth edition of DSSGx UK, gathering young data scientists from all over the world to build software tools for non-profits and government organisations. The only school of its kind in the UK, and one of only a handful globally, DSSGx UK has been running since 2019.     

“We need to make sure that these techniques are not just used by the big corporates, but are also available to NGOs and charities to carry out their good work,” said Professor Branke, DSSGx Director, who has been involved in the programme right from the beginning.  

Tackling deforestation with analytics 

Satyam Suman, 27, applied for the 12-week DSSG programme soon after he joined his MSc Business Analytics course at Warwick Business School in October 2022. 

For the project that would form the basis of his MSc dissertation, he joined three other students in an international team tackling one of the biggest threats facing humanity today: deforestation.  

The focus was to help the UN-REDD programme forecast deforestation in the Brazilian Amazon, identify its key drivers and develop a data visualisation tool to help policymakers fight the loss of the rainforest more effectively.   

“Processing the data, particularly with the huge geographical area involved, was the biggest challenge,” said Satyam.  

But this is where his geoanalytics expertise came in. Having worked previously on projects with the Indian Space Research Organisation on the back of an MSc in Geoinformatics and also using the insights he had gained from the MSc Business Analytics course, he was able to help his team create a digital map that identified the logging hotspots in the Amazon.  

 “Within nine weeks, we had a model that could predict deforestation one year into the future.”  

Overall, during the 12 weeks of the summer fellowship, 16 young data scientists from nine different countries were able to work closely with technical mentors as well as statistics, maths and computer science experts from the University of Warwick and Warwick Business School. 

“I think there is great potential for DSSG in general,” Satyam said. “There are so many opportunities waiting to be explored.” 

Identifying those at risk of being NEET 

Fellow WBS student Mahima Tendulkar, 25, used to help out as a volunteer teacher in rural schools while she was studying electronics and communication engineering at the National Institute of Technology in Goa, India. So, when the opportunity arose on the DSSGx summer fellowship to work on a project supporting young people from low-income backgrounds she jumped at the chance.

“It was just the kind of project that would inspire me,” she said.  

More than 14 per cent of young people between 18 and 24 years old in England, and nearly five per cent of 16-17 years olds have been categorised as NEET (Not in Education, Employment or Training). With the backing of the EY Foundation, the goal of Mahima’s team was to develop a machine learning algorithm that could flag students at risk of NEET early on so that the local authority involved in the project, Bradford City Council, could devise interventions for those in need.  

The team identified all sorts of demographic and educational performance data as well as socio-economic data to develop predictive models for decision making. 

But it was the approach to that data that was all important.  

“This was definitely one of my key takeaways from the WBS course,” said Mahima.  

“It's about learning how to approach a real-world dataset, recognising the challenges in those datasets, and understanding the important information that might be missing from them. After that, it's a matter of taking that data, applying machine learning tools, and embarking on the journey of result interpretation, followed by making that interpretation accessible through a visualisation tool.” 

She added: “We worked very closely with Bradford City Council who, I think, were really impressed by the possibilities of data science.”  

A data project to locate areas of child poverty 

Mahima’s and Satyam’s teams presented their findings at the DSSGx UK Datafest 2023 at WBS London at The Shard along with other project presentations.  

One of those was delivered by Oliver Fiala, of charity Save the Children, who had been a project partner at DSSGx UK 2022. He returned to Datafest this year to explain how the project results had been adopted by his charity.  

The Child Atlas was launched by Save the Children as a data platform on its website in October, using machine learning to provide a granular picture of child poverty around the world.  

One billion children – four in 10 children worldwide – live in multi-dimensional poverty, lacking proper access to healthcare, nutrition, schooling and housing, or even basic clean water and sanitation. But until now their plight has only received patchy coverage from household surveys on a national level. 

The Child Atlas is breaking this down to a more regional level, applying algorithms to data from a whole range of sources including satellite imagery in order to pinpoint poverty hotspots.  

“The Child Atlas is a perfect example of how everything can come together at a DSSGx summer fellowship and the result then applied to a real-word problem,” said Professor Branke. 

 

Professor Juergen Branke teaches Supply Chain Analytics on the MSc Business Analytics degree as well as Analytics for Management on the Executive MBA, Executive MBA (London), Global Online MBA and Global Online MBA (London) courses. 

Learn more about analytics with the four-day Postgraduate Award in Business Analytics for Executives.