Industry problems
Predicting customer experiences in business banking
Banking and finance
For: Absa Bank
Led by: Prof. Riaan de Jongh
Use a raft of variables on historic transactions, banking products used, and demographics to try predict whether a client is satisfied with the bank's offering... find out more
Unturned stones: Astronomy below the survey threshold
Astronomy
For: Inter-University Institute for Data Intensive Astronomy
Led by: Dr. Jonathan Zwart and Prof. Bruce Bassett
Hunt for increasingly complex signals buried in amongst measurement error and other noise in visibility data captured by MeerKAT/KAT7 pathfinders... find out more
Predicting the abundance and distribution of Cape Leopards
Statistical ecology
For: Cape Leopard Trust
Led by Prof. David Borchers
Use images of leopard captured on camera by "camera traps" to estimate how many leopard survive in the Western Cape, with important implications for their conservation... find out more
Collapsing viral next generation sequencing data
Bioinformatics
For: Center for AIDS Research
Led by: Dr Ben Murrell
Next generation genomic sequencing provides huge opportunities for HIV research, if we can just figure out what to look for. Data reduction is the first step in this process.
Can clever feature extraction improve income prediction models?
Retail and marketing
For: Eighty20
Led by Dr. Melvin Varughese
Learn about clever use of features. Use your imagination to help an online marketing company transform a few demographic variables into a powerful prediction of income.