tSNE or t-distributed stochastic neighbourhood embedding is a technique that uses algorithms to reduce N-dimensional data into two dimensions, while still retaining the structure and integrity of the data. The reduction of dimensionality provides greater insights into and enables better visualisation of multi-dimensional data. The algorithms work by randomly initialising positions for each data point in low-dimensional space and then optimising their positions iteratively to better approximate the relationships between the data points and their neighbours in the high-dimensional space. The appropriate place for each data point in the low dimensional space is represented as a probability distribution for each of its neighbours.
The FlowJo™ blog now features a post describing the members of this family of dimensionality-reducing algorithms. It explains the features of these algorithms and also describes how tSNE has become a powerful data analysis tool in flow cytometry.