Collective Action Game Theory has some deep interconnections with information dynamics.

The 18th Chinese Conference on Complex Network (CCCN2022) is on Nov 19-20, 2022.
The introduction site is here.

Zoom meeting on Nov 18, 2022. In Dropbox folder: Study_D1/0学術活動/PJ_皆野町/20221116駒田さん

One of the rewards of switching my website to Jekyll is the ability to support MathJax, which means I can write LaTeX-like equations that get nicely displayed in a web browser, like this one \( \sqrt{\frac{n!}{k!(n-k)!}} \) or this one \( x^2 + y^2 = r^2 \).

CLIP Interrogator is an interesting stable diffusion demo of Text-to-image prompt inversion, i.e., output the image description text by inputting an image. It might be similar to the GAN Inversion.

D1 Middle Term Report

Improve the GPU server security setting by port rearrangement etc, which will be inspected by the university expert team on Nov 21,2022.

I prepared them for the NIS member python study. Useful for data or paper figure preparation. See in my GitHub repo here.

I am proud to mark this speech of Prof.Zhang about his latest work on the Landau-Siegel Zeros Conjecture. See it in Youtube, here.
More detailed explaination in Chinese, here.

Neural Information Squeezer(NIS), an important presentation of causal emergence by Prof.Zhang.

It is the preparation of the paper submission for DHM, HCI International 2023. The title is Policy-Based Reinforcement Learning for Assortative Matching in Human Behavior Modeling.

According to Prof.Jin’s email, I support Haruki ITAI in his graduate essay works concerning causal inference.

To-DO status. Editing this blog. Install LiNGAM in python environmernt and test its basic functions.

Python tools for causal inference. Updated on Oct 2022.

A friendly causal book of Python by Mr.Ogawa. LiNGAM and bayesian network are two important sections to me.

Cause-Effect can be described as DAG(Directed Acyclic Graph, by Judea Pearl). DAG always requires observable data and an inferential path. The former is the big data in our hands. The latter is our problem. Our current solutions are (1) Known path by domain knowledge, (2) Hypothesis, then statistical analysis. However, while these two solutions are out of the question, how do we do?

Prof.Jin instructed to student on his graduate essay. I did assistant suggestions.

Hyperbolic sapce mignt be a key to explore causal relationship in big data.

Corporate with Aobo on one paper, maybe two, for the Special Issue “Causality and Complex Systems”.