In the last few years, Machine Learning has quickly gone from a niche subject to one with significant relevance to many companies and organizations. Across industries ranging from pharmaceuticals and healthcare, to retail and financial services, Machine Learning has become more widely used for solving new business requirements. But just what is Machine Learning and how does it work? Just how do you teach a machine to learn?
Machine Learning at its most basic is the practice of using algorithms to parse large volumes of data, learn from it, and then make a determination or prediction about something in the world. We can work out the probability of certain events occurring in a specific way, and these values changes as more and more events actually happen. This can then affect the likelihood, or probability, of the next event occurring in that way. So rather than hand-coding software routines with a specific set of instructions to accomplish a particular task, the machine is “trained” using large amounts of data and algorithms that give it the ability to learn how to perform the task. This can be expanded to more and more complicated systems with enough data, i.e. enough previous events having happened, then we can make more and more accurate predictions about the likelihood of a future event.
In total, there are three main types of machine learning: supervised, unsupervised and reinforcement learning.
Supervised Learning – A computer is presented with some example inputs and some desired outputs, with the goal being to learn a general rule that maps those inputs to outputs.
Unsupervised Learning – No labels are given to the learning algorithm, leaving it on its own to find structure in its input. Unsupervised learning can be a goal in and of itself (discovering hidden patterns in data) or a means to an end (feature learning).
Reinforcement Learning – A computer program interacts with a dynamic environment in which it must perform a certain goal (such as driving a vehicle or playing a game against an opponent). The program is provided feedback in terms of rewards and punishments as it navigates its problem space.
In this session, all three of these methods will be discussed, along with when and where each type should best be used. We will look at how business software is being developed, using these methodologies, to build Machine Learning software. How good is it, where and when can it be used, and what is the future direction of this field?
We will then take a deeper look at some of the common algorithms that are used to solve Machine Learning problems (such as Logistic Regression, K-means, classification and regression. We will examine how they are grouped together, how they work, what they are best used for, and how you can choose which algorithm(s) best suit the problem you are trying to solve. We will also examine the differences between Artificial Intelligence, Machine Learning and Deep Learning.
Finally, we will look at how today’s businesses are using machine learning in practice to solve real life business problems. From self-driving cars to fraud detection, from drug discover to loyalty programmes, machine learning is changing the way organisations do business today. We will look at some common business use cases and consider what the future holds in the fascinating and rewarding area of Machine Learning.
This session will be an Intermediate level talk. It is geared towards Architects, Data Scientists, Developers, Software Engineers and anyone with an interest in Data Matching.