Machine Learning refers to the application of artificial intelligence where a computer learns from past experiences (often input data) and makes future predictions. As such, machines are able to learn without being explicitly programmed to do so. Machine Learning is a subfield of artificial intelligence and relies on mathematical principles such as Bayesian statistics, which calculates the probability of an event outcome by taking any prior knowledge into account. An example would be what is the probability of rain today given that it did not rain yesterday?
There are 3 main machine learning categories:
- Supervised learning: the training data is known and used for the output we are attempting to predict. The fed back outputs can thus be compared against a reference “known answers” to increase precision. Supervised learning exists either as the classification type (output variable is a category for example “safe” or “unsafe”) or the regression type (output variable is a continuous value “stock market predictions”). Semi-supervised learning relies on training data which includes a few desired outputs.
- Inductive learning is the general theory behind supervised learning, mathematically if we are given an input sample (x) and output samples f(x) the goal is to estimate the f function. Let’s take an example, the face recognition used for smartphone/PC security but also social media. Here, the input x would be the bitmaps of the people’s faces while the aim or function f(x) would be to match the face to owner to unlock the phone or assign a name to the face.
- Unsupervised learning: it differs from supervised learning in that the correct output is not predicted here rather it aims to establish patterns in the input data. Unsupervised models can either be sorted in clusters (grouped based on a specific selection criteria “movie type”) or in association (establishes linkage between different groups, people who buy a product A also tend to buy product B).
Machine learning algorithms usually rely on detailed processes which are often as follows: understanding the prior knowledge and goals, engaging with domain experts, selecting and cleaning data, developing and choosing learning models, analysing and interpreting the results, building up on the obtained knowledge.
Featured Courses
The CPD accredited courses are carefully crafted to help you gain in-depth knowledge on a topic of your interest.
Next Generation Sequencing (NGS)
This intermediate-level course provides a thorough overview of the field of next-generation sequencing, from its properties and strength and its major applications, to an overview of a typical sequencing workflow.
Molecular Microbiology
Molecular microbiology is a very progressive field of science, there is an enormous demand to understand the omnipresent microbes influencing our everyday life on sub-cellular level, which will be subject of this course.
Gene Therapy
Gene therapy is one of the youngest disciplines providing a novel way into treating diseases.
Cancer Biology: Exploring the Molecular and Genetic Aspects
This is an intermediate-level course that explores the molecular and genetic basis of cancer, its evolution, and cancer immunology.
Neuroscience
Neuroscience is one of the most advanced and fastest growing sciences. The beginning of the 21st century has seen huge developments in techniques to study and understand how the brain works.
CRISPR: Revolutionising Genome Editing
This advanced-level course is designed to develop an understanding of complex concepts related to genomics, particularly CRISPR/Cas technology and its applications...
Machine Learning applications are present in our daily lives:
- Bank offers: selectively sent to specific customers for example credit card offers
- Browsing and surfing the web: the displayed pages are ranked based on the ones you are most likely to click on
- E-commerce: predicting what customers are likely to buy based on their search history and preferences but also detect against fraudulent transactions
