Automating machine learning for time-series data

Darko Matovski, CausaLens

Measurements about economic activity are primarily represented in the form of time-series data. This is always the case for real-time data generated from connected devices. Time-series data is unique due to the presence of temporal characteristics. Building models using this type of data is difficult and requires specialised expertise while at the same time there are many algorithms available for analysis and modelling.

Required audience experience

none

Objective of the talk

Help the audience understand how Machine Learning can be fully automated. How a non-technical user can use a Machine Learning platform and extract features associations and predictions at a click of a button while all platforms in the market like IBM Watson and Amazon Alexa need a technical user.

You can view Darko’s presentation below:

Track 1
Location: Auditorium Date: October 16, 2018 Time: 3:35 pm - 4:20 pm Darko Matovski, CausaLens Darko Matovski, causaLens