In our latest 'speaker snapshot' we hear from Richard Mapes, who's close to celebrating his first year anniversary as Director Data Science and Analytics at YPlan.
How did you get started in analytics?
I started out as a neuroscientist but when I moved from academia to the commercial world it never occurred to me that I could take my data analysis skills with me so I went down the software engineer route. Throughout my career, though, I found myself increasingly moving towards data processing and analysis, whether it be using banking data in fraud detection, numerical modeling in algorithmic trading platforms, or Bayesian projection to calculate TV audience demographics from biased samples. I would say I really made the shift to the data science world when I became head of engineering at Qubit, a company specializing in providing an analytics platform and tag management suite for online businesses.
Are there any recent innovations in the analytics data science community that you see as a ‘game changer’?
I wouldn't necessarily say that any one tool or process is a game changer. More that the convergence of machine learning with statistics is the game changer, with experts from the worlds of computer science and mathematics meeting head on and challenging each others assumptions about how things work. We're breeding a new generation and a new field which is neither mathematics nor computers but a sum greater than both those parts.
What are the unique challenges facing YPlan that you are looking to solve with data science?
I wouldn't say unique because I feel we're all facing the same problems - how do I understand what my customer wants and how do I know if we're providing it. I would say one of the biggest problems we have at the moment is keeping track of user identity - mobile devices are becoming actively hostile to tracking and while this is great for avoid targeted advertising, its annoying when you're trying to provide personalized recommendations and the users don't understand why you don't just 'know it's them'.
What will you be discussing in your presentation?
I shall be talking about the importance of data cleansing, analytics pipelines and why the trend for combining your data collection and analytics package into one-third party provider is a bad thing.
Richard Mapes will be presenting at the Predictive Analytics Innovation summit in London on May 11 & 12