Documentation


Viewing posts for the category Omarine User's Manual

I Ching Algorithm: Philosophy and method

In genetic algorithms, a population of rabbits is depicted to illustrate evolution. The rabbits must adapt to the environment, must be smarter and more agile to avoid being eaten by foxes. The ones that survived began to breed, which included smart and agile ones that escaped fox hunting, but there were also some slow and stupid rabbits who survived just for the sake of luck. So different situations happen: some fast rabbits have babies with fast rabbits, some fast rabbits with slow rabbits, some smart rabbits with stupid rabbits, ... etc.
The genetic algorithm implements a completely random reproduction mechanism, which leads to the possibility that the population cannot evolve, the algorithm does not converge to the global optimal. That's because random is not natural. All growth of things have rules.

I Ching Algorithm are intended to replace genetic algorithms and will become completion evolutionary algorithm. Everything has a couple. Boys and girls love each other because of vibration, getting married is through understanding, even if arranged by parents, there is a certain correspondence. It is very rare for a smart person to marry a dull person. Above all, true couple happiness is determined by whether the marriage is good or evil. Nowadays, many people consider boys and girls suitable for their age before deciding to get married. That is the law of the universe. Knowing the destiny of heaven is to find good to avoid evil. Everything is like that, every time in heaven and earth, it is like that.

I Ching Algorithm therefore selects only hexagrams of good to evolve the population, ensuring two goals: fast and correct. Correct evolution means that the result of the problem is the globally optimal solution.


The concepts
The algorithm maintains a population of individuals. Each individual is a vector of elements. The individual is represented by a chromosome, an element that is a gene of a chromosome, in genetics language. Populations evolve by reproduction and selection over generations. The two basic operations for reproduction are chromosomal hybridization and gene mutation (similar to the way the problem is posed in a genetic algorithm, but with a different method).
When the spawning operation meets a good hexagram, the child is born and is included in the population. On the contrary, if the hexagram is evil, it does not apply.
The object hexagram be helpful for the jubject hexagram is good, the objec thexagram and the jubject hexagram are equal is good, the  subject hexagram be harmful for the object hexagram is also good.
The object hexagram be harmful for the jubject hexagram is evil.
The master hexagram be good the variable hexagram be evil means the former is good the later is evil, is evil.
The master hexagram be evil the variable hexagram be good means the former is evil the later is good, is good.
The master hexagram be good the variable hexagram be good means the former is good the later is good also, is very good.
 (if applying probability, the probability of choosing is 100%). Unlike genetic algorithms that only select individuals according to fitness, a dominant individual will become a superindividual leading to a local minima risk.
I Ching Algorithm taking the very good hexagram with 100% probability will improve the fitness of the population while always ensuring the overall picture.


The theorem of problem space representation
Every problem, including binary and numerical optimization, can be represented as binary sequences.

I Ching Algorithm: Illustrate the prediction of the trigrams by divination of name

Once we know how to calculate the number of strokes in the name, we apply it to the case of a girl, whose full name is
PHẠM BĂNG NGỌC
Last name: PHẠM = 2+3+3+4+1 = 13
Padding last name: BĂNG = 3+5+3+3 = 14
First name: NGỌC = 3+3+1+1+1 = 9
Family fortune: 13+14 = 27
Name fortune: 9
Total fortune: 27+9 = 36

I Ching Algorithm: Calculate the number of strokes in the name

Starting from this article we will go to the main part, which is the theoretical basis of the algorithm. The I Ching Algorithm introduced here is the latest evolution algorithm in the global scope and also the first evolutionary algorithm in the 21st century when our species is throbbing with the world of AI.
The core idea of the algorithm is to apply the laws of the universe to the process of reproduction and selection. This introduces general knowledge into evolution ensuring that the algorithm definitely converges instead of running the risk of falling into an infinite loop due to random selection and reproduction as has been demonstrated for classical Genetic Algorithms of the last century.

In the previous articles I could not translate the content into English because there are too many terms describing Eastern philosophy and cosmology. The rest is mainly confined to the technical scope, so I try to be bilingual, but inevitably there are places where the meaning is not clear. In this case, if there are English phrases that are not in the dictionary, please help correct them in the comments or email, otherwise we suggest them as new English terms.

For an introduction to the algorithm, let's look at a familiar problem, which is name prediction. See how a person's name affects the client's life according to the laws of the universe? Analysis of the trigrams in fortune telling is an illustration of all numerical problems using the rule of the trigrams in the I Ching Algorithm.

We know that every abnormal phenomenon has a cause and obeys its laws of motion. What we call Random is something we don't know yet. Obviously what we don't know is huge. However, over thousands of years of history, humans have generalized the law of the universe so that today we can apply it to initialize data with machine learning to solve non-trivial artificial intelligence problems. We don't know when the COVID-19 pandemic will end, so why don't we start from the immutable law "When the rain stops, the sun has to rise".
That is the main meaning of the I Ching Algorithm.

In geometry there are three projections of three-dimensional space. These are vertical view, side view and flat view. Any projection carries information about the shape of the object. Roughly, so is the law of the universe, each phenomenon reflects an aspect of the nature of things. But unlike physical phenomena, the universe is represented in an infinite dimensional space that is not linearly independent. That means we either don't see it, or just an induciable root information phenomenon (refer to decision tree) can cover the whole problem.

KINH DỊCH algorithm: Mười thần

Các thiên can được ánh xạ vào 10 thần để mô tả các vấn đề xã hội, lý tưởng, tiền tài, công việc, sự nghiệp và lục thân.
Mười thần là Tỷ kiên, Kiếp tài, Thực thần, Thương quan, Thiên tài, Chính tài, Thiên quan, Chính quan, Thiên ấn, Chính ấn.
Những thần có từ Thiên ở trong tên là không chính thống, có từ Chính là chính thức, chính thống. Ví dụ Thiên ấn là mẹ nuôi, mẹ kế, Chính ấn là mẹ đẻ.
Nếu Thiên quan có hại, tấn công thân, là tiểu nhân hãm hại mà không bị khắc chế thì được gọi là Thất sát. Bản thân các thần là thiên can nên có thuộc tính ngũ hành. Nếu nhật chủ là Giáp mộc thì sẽ bị kim khắc. Dương khắc dương, âm khắc âm là Thiên quan. Dương khắc âm, âm khắc dương là Chính quan. Do đó Thiên quan sẽ là Canh dương kim. Đếm từ Giáp đến Canh là 7 nên gọi là Thất sát.
Nếu Thiên ấn là bà mẹ ghẻ độc ác chuyên đoạt Thực thần thì được gọi là Kiêu thần.
Các thần được gọi tắt là tỉ, kiếp, thực, thương, tài, sát, quan, kiêu, ấn.

KINH DỊCH algorithm: Trong tứ trụ, trụ nào quan trọng nhất?

Trên cơ sở lý luận trụ ngày quan trọng nhất vì nó là nhật chủ, là thân chủ. Thứ hai là trụ tháng vì lệnh tháng giữ quyền sinh quyền sát đối với các thiên can. Nó quyết định nhật chủ cường vượng hay suy nhược. Lệnh tháng cũng quy định trạng thái sinh, vượng, tử, tuyệt của các can khác.
Trụ năm là gốc, mang thông tin về cha mẹ tổ tiên và dòng dõi xuất xứ của một người cũng như thời thơ ấu, quan trọng thứ ba.
Cuối cùng là trụ giờ, là thông tin về con cái và lúc về già.

Tuy nhiên, chương trình ./fpp --id3 mang luận cứ khoa học vững chắc về xác định lượng thông tin và xây dựng cây quyết định, đã đưa ra yếu tố quan trọng nhất (root) là chi tháng (tức lệnh tháng) chứ không phải là can ngày (nhật chủ) sau khi duyệt qua 202 mẫu như hình minh họa dưới đây