Open source forecasting with Python and Prophet
errazudin ishak (~errazudin) |
Facebook Prophet claimed to produce forecasts that are often as precise as those produced by seasoned forecasters, with much less effort and just using default settings? This talk aim to uncover the truth(read:pain) with potentially some working examples on real-life case.
Forecasting at scale using one of the open source projects from Facebook. Prophet claimed to produce forecasts that are often as precise as those produced by seasoned forecasters, with much less effort and just using default settings. Sounds great? This talk tries to uncover some of the love (and hate) points when dealing with this tool and some real-life case studies.
Python, Jupyter, Time series data, Data analysis
Errazudin works as Solutions Architect at MTSSB, Malaysia. His job focuses on building software, architecting solutions, cloud wrangling and making sense out of massive data. He has previously spoken at several meetups and conferences and has worked with various technologies. In his free time, he loves to compose Ahmad-Jais-like tunes in his dream or try to help his little daughter with her super tough kindergarten homework.
https://www.linkedin.com/in/errazudinishak/ https://www.slideshare.net/errazudin https://speakerdeck.com/errazudin