Data Driven London
Wednesday, April 24 at 6:00 PM
We are celebrating Big Data Week in April! This event is for entrepreneurs and professionals interested in real time metrics, lean analytics and d…
Bigdataweek.com #bdw13 #datadrivenlondon @geckoboard
This event is for entrepreneurs and professionals interested in real time metrics, lean analytics and data driven decision making.
Mark RobinsonCOO GamesAnalytics
Putting Player Experience at the heart of Game Development
Fascinated by the potential of big data and games, Mark co-founded GamesAnalytics in 2010 and has made it his personal mission to evangelise how analytics can change the games industry learning CRM techniques and business thinking from other market sectors.
Mark Robinson has over 15 years’ data mining experience industry, across companies such as Heineken, Office Depot, Aviva and Unibet. He led data mining consultancy Marketing Databasics that builds and hosts analytical environments to increase customer value through data-driven insight. In 2009, Marketing Databasics (MDB) was designated a “Leader” organisation in the Forrester review of marketing services providers for its completeness of offering and its strategic vision. MDB was acquired by the Indicia Group where Mark was Client Services Director before leaving to start GamesAnalytics as COO.
GamesAnalytics to enable big data to drive player understanding, introducing the concept of Player Relationship Management to build better games. With the advent of the Free-2-Play games, responding to player behaviours has never been so important and GamesAnalytics helps publishers & developers with player segmentation, predictive analytics and targeted in-game messaging in order to increase player engagement and game revenues.
Lean Predictive Analytics
Thomas is Co-founder of PredictionIO and a PhD candidate in Computer Science at University College London with a passion for entrepreneurship and data analytics. Thomas brings prior startup experience plus knowledge of early-stage companies and venture finance having previously worked with companies such as Silicon Valley Bank, Seedcamp and UCL Advances. He graduated from University of Birmingham with First Class honors, studied abroad at Cornell University and has a Masters in CS from University College London.
PredictionIO is an open-source machine learning server.
PredictionIO enables developers and data engineers to build smarter apps. Developers can add predictive features to their web or mobile applications easily through a simple set of APIs. For instance, they can provide product recommendations in the way Amazon does, or news recommendations in the same fashion as Flipboard, with just a few lines of code. For more information see http://prediction.io/
We will be hosting a panel of experts and will be asking key questions around:
Big Data Vs Data Communication
How to build and leverage data driven culture?
How to make data approachable, visible and actionable?
Speakers and panelists to be announced next week.
This event will take place at Google Campus, 3rd Floor.
If you have any questions please get in touch: [masked]
Game analytics : http://www.gameanalytics.com responding directly to gamers for the first time.
Getting inside the minds of the customer. Success, momentum, frustrations and improve the experience. These processes actually come from outside the game industry.
Engage (personalised messages) boost, (retention, segmentation, metrics) measure (performance metrics and data collection) and benchmark. (Best practice design)
Sporadic semi engaged
Game analytics is a fundamental change in the industry of games is a game changer. (Sound familiar BIM readers? There was me thinking this was commonplace practice)
Thomas Stone – PredictionIO
Aimed at people who understand API, SDK and writing apps.
Machine Learning http://blog.prediction.io/machine-learning/
PredictionIO Open Source Machine Learning Server
PredictionIO is an open source prediction server for software developers to create predictive features, such as personalization, recommendation and content …
Predictive Analytics : finance and amazon/lovefilm for making suggestions
Lean predictive analytics (?)
Computers learning to predict from data
Challenge #1 Learning Curve – k-nearest neighbour (kNN) and weight sums. Collaborative filtering.
Challenge #2 Scalability , Big Data Bottlenecks – horizontally scalable architecture.
Use Cases :
Important to funnel results of analysis back into the business to affect outcomes.
There are parallels between optimising gamers experience to that of building outcomes for a building user/manager.
How much data is enough data? Depends on the prediction goal. Requires users, selections and ratings.
Management of data is driven by the data protection laws
Creating open source and then allowing the client to store the data.
Analyse, develop, monitor. Loop of this process. 10,000 data points from 1,000 users as a starting point.
Real time insights where decisions need to be made in a shorter time frame. To be able to provide promotion of relevant data by popularity. everything needs to happen now and react now.
Data collection is key and is often the limitation, how can we in the construction industry capture data and is it fast enough to allow us to react? Real time analysis and response is going to be a game changer in lean processes and businesses who are early adapters are going to gain the competitive advantage.