Have the right data for the analytics solution you are building. John Kellegher, DIT

By @SimonCocking
John Kellegher, DIT, speaking at the Predict Conference September 15 – 17th 

How did you first become interested in data / predictive analytics?

I first became interested in data/predictive anaytics through teaching machine learning modules at the Dublin Institute of Technology.

What is your best data (data modelling/predictive analysis) tip?

‘Garbage in Garbage out’: without good data your project won’t be successful.

What advice would you give to someone just starting out on their data journey?

(1) Make sure you have the right data for the analytics solution you are trying to build.
(2) Spend time exploring different models and different parameter settings.
(3) Make sure you are rigorous in designing and implementing your evaluation experiments.

What skills do you think a good data scientist analyst should have?

Having a numeric aptitude is obviously very useful. However, the ability to listen and to communicate is also incredibly important. If a data analyst isn’t able to listen to and learn from business focused collegues then the analyst won’t be able to understand the business needs of a project and as a result the project is not likely to succeed. Conversely, unless a data analyst is able to communicate how a model works, the assumptions it makes, and the accuracy of the model, then the business won’t be able to leverage the insight from the model appropriately.

What resources would you recommend (e.g. books, websites, blogs, data, technology, etc.)?

I have just co-authored a book on machine learning and data analytics called:

‘Fundamentals of Machine Learning for Predictive Data Analytics: Algorithms, Worked Examples, and Case Studies’ published by MIT Press.

Obviously, I am slightly biased in my evaluation of the book but I do believe that it is an excellent book that explains the key machine learning algorithms most people will end up using in business and, perhaps more importantly, explains how machine learning fits into the broader data analytics lifecycle within a business. See here.

More about John Kellegher below

John is a lecturer at the School of Computing in Dublin Institute of Technology (DIT) and the manager of the Applied Intelligence Research Centre at DIT. He is currently a collaborative researcher at CeADAR  and a funded researcher in the Adapt research centre. Before joing DIT, John had worked as a researcher at a number of institutes including Media Lab Europe and the German Center for Artificial Intelligence (DFKI). During his research career John has worked on a range of topics from robotics to text analytics. Machine learning and data analytics has been a constant across these projects. John has written over 80 peer reviewed publications and is co-author of the text book Fundamentals of Machine Learning for Predictive Data Analytics: Algorithms, Worked Examples, and Case Studies published by MIT Press in 2015.

You can see John and many more speak at the Predict Conference in Ireland at the RDS, DUBLIN, IRELAND 15 – 17 SEPTEMBER 2015