Major research fields
computational neuroscience, brain machine interface, data science
Although the brain processes information via the electric pulses exchanged among neurons, it is not easy to understand the algorithms for its purposeful computations as a whole only by observing individual neurons experimentally. Therefore our laboratory, as a dry lab, uses computers and theoretical models to understand the brain.
One of the goals of computational neuroscience is to crack the "neural codes" of the electric pulses enough to read animals’ minds. For example, now it is possible to decode the odors a rat smells from the measured neurons’ electric activities by using machine learning methods. Here you can decode more accurately by hypothesizing that not only the frequencies, but also the timings of electric pulses are important. In this way, you can obtain a better understanding of the neural codes via various hypotheses. In addition to senses including odors, more recently, researchers began to decipher higher brain functions such as “motivations” and “values”. Here the roles of hypotheses and theoretical models are essential. In a dry lab, the ideas for mining valuable information from big data are essential. If you learn data science skills during your graduation work, which is actually pretty easy with packaged software, the skills will also work for business data analyses.
Major relevant publications
- Miura K, Mainen ZF, Uchida N, Odor representations in olfactory cortex: distributed rate coding and decorrelated population activity. Neuron, 74, 1087-1098, (2012).
- Wang AY, Miura K, Uchida N, The dorsomedial striatum encodes net expected return, critical for energizing performance vigor. Nature Neuroscience, 16, 639-647, (2013).
- Miura K, A Semiparametric Covariance Estimator Immune to Arbitrary Signal Drift. Interdisciplinary Information Sciences, 19(1), 35-41, (2013).
- Miura K, Nakada K., Neural Implementation of Shape-Invariant Touch Counter Based on Euler Calculus. IEEE Access, 2, 960-970, (2014).
- Miura K, Aoki T, Hodge-Kodaira Decomposition of Evolving Neural Networks. Neural Networks, 62, 20-24, (2015).