Neural Information Processing System

Course title

6M019000

Course content

We have a brain and nervous system that processes information obtained from our sensory organs to confront the ever-changing environment. We also respond to dynamic environments by continuously rewriting our neural systems through learning. In this lecture, we will learn the methodology to understand the mechanism of the cranial nervous system from a computational approach.

Purpose of class

The purpose of this lecture is to understand basic theories of the nervous system, especially neural networks, reinforcement learning, and Bayesian inference.

Goals and objectives

Can explain basic concepts and fundamental theories of neural networks
Can explain basic concepts and fundamental theories of reinforcement learning
Can explain basic concepts and fundamental theories of Bayesian inference

Language

Japanese

Class schedule

	Class schedule	HW assignments (Including preparation and review of the class.)	Amount of Time Required
1．	Introduction: Computational neuroscience	Review of Introduction to Computational Neuroscience	60minutes
2．	computational approach	Review of computational approaches	200minutes
3．	Biophysical Models (1) Overview of the nervous system and neurons	Review of neurons	200minutes
4．	Biophysical Models (2) Description of typical biophysical models	Review of biophysical models	200minutes
5．	Neural Networks (1) Principles of operation and learning of neural network models	Review of Neural Networks	200minutes
6．	Neural Networks (2) Recurrent neural network	Review of Recurrent Neural Networks	200minutes
7．	Neural Networks (3) Initial Value Sensitivity, Top-down Prediction, and Bottom-up Correction	Review of initial value sensitivities, predictions and corrections	200minutes
8．	Reinforcement Learning Model (1) Reinforcement Learning Model Based on Behavioral Value	Review of Behavioral Value Dependent Modeling	200minutes
9．	Reinforcement Learning Model (2) Reinforcement Learning Model Based on State-Values	Review of state-value dependent modeling	200minutes
10．	Reinforcement Learning Models (3) Reinforcement learning model dealing with state transitions and delayed rewards	Review of Reinforcement Learning Models	200minutes
11．	Bayesian Inference Model (1) Bayesian inference	Review of Bayesian Inference	200minutes
12．	Bayesian inference model (2) Kalman filter	Review of Kalman filter	200minutes
13．	Bayesian Inference Model (3) Hierarchical Gaussian filter	Review of Hierarchical Gaussian Filters	200minutes
14．	Bayesian inference model (4) free energy principle	Review of free energy principles	200minutes
Total.	-	-	2660minutes

Relationship between 'Goals and Objectives' and 'Course Outcomes'

	レポート	プレゼンテーション	Total.
1.	30%	0%	30%
2.	30%		30%
3.	40%		40%
Total.	100%	0%	-

Evaluation method and criteria

A score of 60 or higher on the submitted report will be considered acceptable.
A score of 60 is equivalent to a simple explanation of the basic theories of the nervous system and computational neuroscience.

Textbooks and reference materials

Handouts will be distributed as needed.
Reference book: "Computational Psychiatry" by Aihiko Kunisato, Kentaro Katahira, Sai Okimura, Yuichi Yamashita, Keiso Shobo, 2019 (in Japanese)

Prerequisites

Nothing in particular, but think about your own relationship to neuroscience.

Office hours and How to contact professors for questions

Tuesday 12:30-13:20

Regionally-oriented

Non-regionally-oriented course

Development of social and professional independence

Non-social and professional independence development course

Active-learning course

N/A

Course by professor with work experience

Work experience	Work experience and relevance to the course content if applicable
N/A	not applicable

Education related SDGs:the Sustainable Development Goals

3.GOOD HEALTH AND WELL-BEING
9.INDUSTRY, INNOVATION AND INFRASTRUCTURE

Last modified : Sat Mar 19 00:42:35 JST 2022