Python September Meetup

# Talks

1. Deep learning for time series by Alexander Honchar (@alexrachnog) (English, 60 min)

10 years ago software was eating the world, today AI is eating the software. Machine learning is developing extremely quickly, especially such fields as computer vision, natural language processing, business oriented fields as recommendation systems and many others. In shadow of hyped areas like computer vision and NLP we can forget about another important source of data - signals, or time series.

In this talk I want to review the world of applications of time series analysis (since application areas are not very known to the broad audience), discuss previous approaches (let’s call it “pre-deep-learning era”) and share my experience in using deep neural networks in bio signal analysis and financial forecasting. Apart of best practices I’ll show some important research paths that potentially can solve any time series related problems in the future.