Simon Chan

PredictionIO

Bio

Simon is the CEO and co-founder of PredictionIO - an open source Machine Learning Server that empowers programmers and data engineers to build smart applications. Starting off as a software engineer, Simon founded three tech startups in the past 10 years, in the bay area, in Hong Kong and in Mainland China. He specializes in machine learning and recommendation technology, with a strong interest in social applications. Simon holds a B.S.E. in Computer Science from University of Michigan, Ann Arbor and is a PhD candidate in Machine Learning at University College London.

Anticipatory Computing, Smart Devices and Machine Learning

Room: N-125 | Time: 2:30pm - 2:50pm

PredictionIO is an open source machine learning server for software developers to create predictive features. Traditionally, this included personalization, recommendation and content discovery in domains such as fashion, e-health and gaming. The latest version of PredictionIO will open our platform for many more use cases such as churn analysis, trend detection and event anticipation. Allowing developers to use the power of machine learning not only for web and mobile apps but smart objects and connected devices. In this talk we will give an application of event anticipation based on patterns of behavior suggesting the next appropriate action e.g., 'Turn down the lights?' based on learning from past actions. We will also discuss the new software design pattern for building machine learning engines in Scala, R and Python on top of PredictionIO's scalable infrastructure. In recent years products like Google Now and MindMeld have pioneered anticipatory computing. Now it's time to see what an open source community can build re-imaging the objects and devices that surround us in our daily lives using machine learning.

2:30pm - 2:50pm