Jukka Aakula > Recent Status Updates

Data normalization perspective to PCA.

Erinomainen, hauska, traaginen.

Huh huh. Yksityiskohta yksityiskohtaa yksityiskohtaa

Aivan liikaa yksityiskohtia.

Canetti matkusti Marrakeshiin 1950-luvulla, ja matkan pohjalta syntynyt teos julkaistiin 1967. Kirja ei ole ehkä varsinainen matkakertomus vaan jonkinlainen tutkielma.

Kaunis ja viisas kirja.

The writer has held lectures on this subject to undergraduates, every time improving his material.

And it has become a beautiful book. Different from a typical theoretical math book, where you first get the theory, and then you can work on the applications. On page ~40, you already know the idea quite well and what to do with the PCA.

But not an easy book.

From intuition to a more formal part.

The first chapter shows the idea in a way that is not just an intuitive introduction, but also shows a crude solution to the problem at hand.

But it is not an easy introduction. Also, you should have some understanding of matrix calculus.

"In an open world economy, it is almost always cheaper to trade for a particular natural resource or industrial product than to take it by force. But with foreign markets locked away behind insurmountable tariff walls, conquest once again became thinkable and even profitable."

Think about today.

The failure of French revolution and the success of the organic "revolutions" (that is, cultural and institutional evolution) of the Dutch and British lands.

Excellent at least so far.

Big Goods theory is wrong according to the Seshat analysis.

Chapter 7.6 was partly too difficult for me to understand, even if the threshold model of Collective computation and Scale is clear. I skipped part of the text.

90% of the book is accessible with basic knowledge on statistics, cultural evolution, multilevel selection, agent-based models, etc.

Military technologies and agricultural intensity are the evolutionary drivers for complex societies. Other drivers are not very relevant. Also the affect is pretty much unidirectional. E.g. military technology (until 1500) has been external cause.

"Can different human societies, from different parts of the globe and from different historical eras, be meaningfully compared in their complexity? And how many measures do we need for this comparison: one, two, or many?

The answers to these questions are “yes” and “one,” respectively. Furthermore, we now have a quantitative answer to what is meant by the “bulk of variation”—79 percent, captured by PC1."

"The synthetic theory (Chapter 3) postulates that technical advances in the military sphere increase the intensity of interpolity competition, thus selecting for larger-scale, better-organized, and more cohesive societies. The Seshat project has developed a sophisticated approach for quantifying this potential driver of social evolution (Turchin et al. 2021c)."

Dynamic regression is used to analyze dynamism.

Chapter 3 provides the theoretical synthesis for the Great Holocene Transformation. Probably the main non-formal argument of the book is now defined.