DSPT#56 – How to forecast at scale and pull out the big GANs (Aveiro)
Ranging from scaled time series prediction, showing some problem scenarios from small to large clients, to GAN (Generative Adversarial Networks), where you can learn useful representation or use it to falsify things.
On the 4th of June at Ubiwhere, we hosted DSPT’s 56th. This evening, we had two amazing talks with André Antunes, from SCUBIC, and Eduardo Pinho, a Ph.D. student in computer science at IEETA.
In the first talk, André Antunes shared with us how we can move a small client to 2 of the 3 largest clients in the Portuguese market, ensuring that the same model creation procedure can be applied to different scenarios, in different geographical locations, to different demographics, faster to ensure the same performance.
In the second talk, Eduardo Pinho introduced us to how GANs can be valuable in using unsupervised learning methods using the use case of medical image information extraction.
Finally, we would like to thank Ubiwhere for cooperating with DSPT in organizing this meetup.
And of course, no DSPT meetup would be complete without our group photo.
The event in numbers
- 2 Speaker
- 1 Coffee break
- Lots of minis
- 2.5 hours of shared knowledge
Interesting links and contents related to the meetup.