Dimensionality reduction for large-scale neural recordings

By John P. Cunningham, Byron M. Yu. 2014

1,240 citations recorded by the source corpus.

Most sensory, cognitive and motor functions depend on the interactions of many neurons. In recent years, there has been rapid development and increasing use of technologies for recording from large numbers of neurons, either sequentially or simultaneously. A key question is what scientific insight can be gained by studying a population of recorded neurons beyond studying each neuron individually. Here, we examine three important motivations for population studies: single-trial hypotheses requiring statistical power, hypotheses of population response structure and exploratory analyses of large data sets. Many recent studies have adopted dimensionality reduction to analyze these populations and to find features that are not apparent at the level of individual neurons. We describe the dimensionality reduction methods commonly applied to population activity and offer practical advice about selecting methods and interpreting their outputs. This review is intended for experimental and computational researchers who seek to understand the role dimensionality reduction has had and can have in systems neuroscience, and who seek to apply these methods to their own data.

Subfields: Artificial Intelligence, Cellular and Molecular Neuroscience, Cognitive Neuroscience

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cellular and Molecular Neuroscience

In Cognitive Neuroscience

Citations in the corpus

Re-evaluating circuit mechanisms underlying pattern separation (2019, neutral): “In principle, measuring dimensionality is as straightforward as recording a population of neurons and using principal component analysis (or a nonlinear dimensionality reduction method) to determine how many components are necessary to capture the majority of the variance (Cunningham and Yu, 2016).”
Computation Through Neural Population Dynamics (2020, background): “Many studies have reported that neural population dynamics tend to lie within a low-dimensional subspace (or manifold) embedded within the high-dimensional neural population state space (Cunningham & Yu 2014).”
Scalable Bayesian GPFA with automatic relevance determination and discrete noise models (2021, methodology): “We proceeded to fit both factor analysis (FA), GPFA, and bGPFA with different latent dimensionalities D ∈ [1, 10].”
Interpreting neural computations by examining intrinsic and embedding dimensionality of neural activity (2021, background): “While a number of algorithms have been developed for non-linear dimensionality reduction, accurately estimating intrinsic and embedding dimensions from noisy data remains an active area of research [22,26].”
Complementary roles of dimensionality and population structure in neural computations (2020, methodology): “In line with dimensionality reduction approaches [Cunningham and Yu, 2014; Gallego et al., 2017], in this framework, the network connectivity is represented in terms of patterns over neurons (Fig.”
Cortical areas interact through a communication subspace (2019, positive): “, 2016), as identified using dimensionality reduction (Cunningham and Yu, 2014).”
Unsupervised Discovery of Demixed, Low-Dimensional Neural Dynamics across Multiple Timescales through Tensor Component Analysis (2017, methodology): “Neuroscientists have often turned to unbiased dimensionality reduction methods to understand these complex 27 datasets [12, 13].”
Network neuroscience (2017, neutral): “The size and complexity of these data require sophisticated strategies for statistical inference and dimensionality reduction 7 and present important challenges for scalable computing 8 , data sharing, and reproducibility 9 .”
Projections of non-invasive human recordings into state space show unfolding of spontaneous and over-trained choice (2020, methodology): “29ms+/-10ms; effect of training on % correct: F(1,20)=36.”
Computational psychiatry as a bridge from neuroscience to clinical applications (2016, background): “For example, models of a variety of time-varying processes, such as learning 22 , multi-neuron recordings 23 and BOLD time series 24 , can collapse long and seemingly complex time series into a few parameters characterizing the underlying dynamics.”
Neural population coding: combining insights from microscopic and mass signals (2015, methodology): “Progress in understanding state-dependent coding requires better statistical methods for single-trial analysis that can be applied to different measurements of brain activity to disentangle state-dependent and -independent aspects of a code [12,79–81].”
Linear dimensionality reduction: survey, insights, and generalizations (2014, methodology): “Indeed, such methods have been invented with slight variations in various application areas (e.g., Cunningham and Yu (2014)).”
On simplicity and complexity in the brave new world of large-scale neuroscience. (2015, neutral): “-Albert Einstein.”