![]() ![]() There are many different metrics to calculate distance. We can compute the representational space of these stimuli by calculating the pairwise distance between each image in this feature space (e.g., euclidean, correlation, cosine, etc). What if were were interested in more low-level visual features? Of course, this is a just a 2-dimensional example, this idea can be extended to n-dimensions. Where would a puppy, rock, and pillow be positioned within this two dimensional embedding space? For example, consider the two different dimensions of animacy and softness. Think of each feature as an axis in a multidimensional space. We can group together different variables into a feature space and then describe each image using this space. These variables could be more abstract, such as is it animate or inanimate? or they can be more low level, such as what color is each object? Feature embedding space #Įach image can be described by a number of different variables. Do we categorize these images into different groups? If so, how do we do this? It might depend on what features, or aspects, of the stimuli that we consider. Imagine that you view 96 different images in the scanner and we are interested in learning more about how the brain processes information about these stimuli. Let’s work through a quick example to outline the specific steps involved in RSA to make this more concrete. How to map variations in position within this embedding space onto brain processes.įor a more in depth tutorial, we recommend reading this excellent tutorial written by Mark Thornton for the 2018 MIND Summer School, or watching his video walkthrough. How to compute similarity or distance within this feature space How to embed a stimuli in a feature space This sounds complicated, but is actually quite conceptually simple. ![]() Unlike multivariate prediction/classification, RSA does not attempt to directly map brain activity onto a measure, but instead compares the similarities between brain activity and the measure using second-order isomorphisms. This technique was initially proposed by Nikolaus Kriegskorte in 2008. Representational Similarity Analysis (RSA) is a multivariate technique that allows one to link disparate types of data based on shared structure in their similarity (or distance) matrices.
0 Comments
Leave a Reply. |