Machine learning refers to algorithms and techniques to learn from Big Data. It is used when there is not a direct mathematical formula to describe the data and there is an assumption that some kind of pattern exists. In other words we need learning in those cases where we can not directly write a computer program to solve a given problem and instead need to use example data. Above all those algorithms and techniques thus require large quantities of data to learn from. One goal of it is to program algorithms to use example data or past experience to solve a given problem. The techniques can be categorized in supervised learning (e.g. classification tasks based on labelled data), unsupervised learning (e.g. clustering or association rule mining), or reinforcement learning (e.g. game playing moves).
One example application is the prediction what new movies a viewer might like based on historical data of already watched and rated movies. Machine learning algorithms are thus able to recommend viewers certain unrated movies based on their content or actors. Other successful applications are systems that analyze past sales data to predict customer behavior. Other applications focus on the recognition of faces or spoken speech. Yet another application field extracts knowledge from bioinformatics data. Machine learning problems and solutions take advantage of approaches from different fields such as statistcs, pattern recognition, artificial intelligence, signal processing, or data mining. The goal of building learning systems that can adapt to their environment and learn from their experience has attracted researchers from a wide variety of research fields including computer science, engineering, mathematics, physics, neuroscience, and cognitive sciences.
Machine Learning Details
The following LinkedIn tech talk video of Ron Bekkerman provides a good overview of the Topic:
Follow us on Facebook: