Title: Scalable computation for large-scale neural data analysis
Abstract: Advances in neurotechnology have allowed us to record large-scale, high-throughput neural data through electrophysiology and optical imaging. However, current exponential growth in the scale of data acquisition is a double-edged sword. These "big data" present a challenge for various neural data analyses such as decoding and functional connectivity analysis, as well as closed-loop brain-machine interface (BMI) applications in neuroscience experiments. We will discuss the strategies to improve the scalability in systems neuroscience and practical methods to readout large-scale neural recordings in electrophysiology and calcium imaging. We will use the rodent hippocampus as an illustration example and present a few case studies to extract information from large-scale multielectrode arrays or calcium imaging based on unsorted spikes, local field potentials (LFPs) and raw fluorescence traces. Specifically, we show how to apply supervised and unsupervised machine learning, including dimensionality reduction and lower-rank factorization, to various neural neural data analyses. In the end, we will also discuss the trend and opportunity of deep learning in large-scale neural data analysis.
