AN INTRODUCTION TO ACTIVE LEARNING, AND ITS APPLICATION TO BUSINESS COACHING
Speaker: Tom Osborn Seminar Date: Tuesday November 21 12:00pm Brief abstract: Active learning is a branch of semi-supervised machine learning for domains where data is expensive and where precise understanding is critical. Traditionally, active learning builds a model of a domain with a trade-off between exploration and exploitation probing of a domain to generate labelled data usefully. Earlier applications of active learning included intelligent tutoring and robotic search of a novel environment. More recent examples include algorithmic trading in online auctions and markets, recommender systems and fast optimal classifiers. Working with two experienced business coaches, we applied active learning to business coaching. Specifically, our system efficiently diagnoses strategic management issues of a business, by adaptively serving up relevant questions (from a universe of potential questions) to senior executives. The system builds a hierarchical model of business health and risks in an information optimal way. The system also advises and prescribes businesses based on these diagnoses. Essentially the system is “small data”, discovers strategic management gaps which senior execs were unaware of (and confirms those they are aware of). In brief, Longitude Systems’ active learning provides a voice in the room for improving strategic management. Short Bio: Tom is a freelance analytics consultant in predictive modelling, machine learning, computer forensic and strategic analysis. He studied mathematical modelling at the University of Newcastle, finished with a PhD modelling complex systems. He then lectured at UTS in the 1980s and 1990s, in machine learning, AI, simulation, software engineering and formal methods before moving to consulting. Among other roles, Tom was Head of AI/Modelling at NUIX, Chief Scientist at Brandscreen and Data Strategist at Y&R/Wunderman. More recently he is non-executive director and head adviser on analytics with Digivizer, Adjunct Professor (renewal pending) at UTS and R&D partner at Longitude Systems. Longitude is his sixth start up.