Demystifying Machine Learning

5 December 2017, 2.00 PM - 5 December 2017, 4.30 PM

Professor Peter Flach and Dr Niall Twomey

Stephenson Room, Richmond Building, University of Bristol

Machine Learning

Machine learning is frequently in the news these days and is seen as an important enabling technology for artificial intelligence. Patterns detected by machine learning algorithms are being used for machines to make autonomous predictions and decisions. But what is this exciting technology, and how is it different from other, more conventional methods of data analysis? Could it be that you are already doing machine learning, without realising it?

Shedding some light on current machine learning practice

The aim of this session is to shed some light on current machine learning practice. Rather than attempting a precise definition of the field and how it differs from other approaches (or not!), we will show what a machine learning practitioner has in his or her toolbox for day-to-day use. Chances are that you will be at least somewhat familiar with some of these tools, such as linear regression or k-means clustering. However, other approaches may be less familiar, such as random forests or support vector machines.

Getting the data in right format

It is sometimes said that building machine learning models is only a small part - perhaps 10% - of a successful machine learning project. The other 90% is spent on getting the data in the right format and tuning the modelling tools to the problem at hand. In the second part of the session we will take a closer look at machine learning-based data science in practice, addressing topics such as data ingress and pre-processing, data storage and management, and data transformation and integration.

Is this workshop suitable for me?

The first part should be accessible for a general audience with a basic grasp of mathematical notation and some knowledge of basic techniques such as linear regression. The second part will require some knowledge of computer programming and involve live demonstrations which people can replicate on their laptops if they choose. Depending on their background attendees may choose to attend both parts or only the first part.

Contact information

To register for this workshop please visit Demystifying Machine Learning

Edit this page