Mind blowing ways to solve the seemingly impossible with analytics
Wednesday, 19th August 2015
An overview of IAPA's Advancing Analytics 2015 National Conference By Pia Harrison @Pia_Harrison1 #advancinganalytics
It was an action packed 3 day event that saw the brightest minds in analytics meet for the annual IAPA Conference in Sydney from 3-5 August 2015. From helping dairy and agriculture farmers to feed the world more efficiently to detecting deception in company annual reports; these problem are no longer insurmountable. Speaker after speaker we heard the creative ways that analytics is being used to see things that on the surface seem completely unrelated and solve some of the most challenging problems.
Starting with challenges that are literally out of this World, Bob Jones from CERN (@BobJonesCERN), described how they use analytics to understand the Universe. They generate the equivalent of 250 years of high definition movies every year worth of streamed data in order to find 400 occurrences of the same event out of 600 million billion observations; illustrating the true volume and veracity of big data. View Bob's presentation here.
Back on planet Earth, Usama Fayyad (@usamaf) from Barclays Bank, went on to describe how to tackle the variety of data in order to regain old fashioned service and customer intimacy at scale. As Usama explained, “structured data is not big data”. So when you are faced with a variety of data sources from social media to video, blogs and publications, how do you store and make sense of it all? View Usama's presentation here.
Speaking of making sense of things, Professor Stephen Pulman (@sgpulman) from Oxford University gave us an insight into training computers to understand emotion and detect deception in written or textual data. With incredible applications including detecting deception in company annual reports to politician’s speeches and border control. View Stephen's presentation here.
If your mind hasn’t already imploded, next up Arkady Zaslavsky from the CSIRO provided an insight into how the Internet of Things is being used to improve agricultural efficiency. Imagine plant sensors that send you a message when they need watering. He also posed the question, “what if all the computers in smart cars parked in the city would be turned into cloud computing and rented out?” View Arkardy's presentation here.
Following on the food production theme, Kevin Ross from Fonterra explained how to optimise milk production. Even cows are wearing fitbits these days in order for farmers to identify when cows are most fertile and ready for conception.View Kevin's presentation here.
An expert panel comprising of Dr Leif Evensen from Westpac, Julie Batch (@JulesBatch) from IAG and Liz Moore (@ms_lizzie) from Telstra discussed what makes the best analytics team? A healthy dose of curiosity and creativity were mentioned; however consensus was that business acumen and communication skills as well as technical ability were essential for high performing teams.
Next up, how to use data to recognise robot generated click-throughs / traffic by Claudia Perlich (@claudia_perlich) from Dstillery. Using data to uncover all sorts of online traffic myths and understand the online path to purchase. Interestingly Claudia explained that it’s “helpful for data scientists not to know too much about the domain.” View Claudia's presentation here.
To top off a stellar day of speakers, Igor Elbert (@ielbert) from Gilt.com online fashion website gave an insight into pre-emptive shipping by predicting what clothing you’ll buy before you buy it. Igor conquered the difficult feat of modeling a creative industry such as fashion by using people via the Mechanical Turk to reduce ambiguity.View Igor's presentation here.
A sensational IAPA conference that is sure to incubate in the minds of all those who attended and hopefully spark new ideas and solve other seemingly impossible problems with analytics.