Of course, Chang Haonan also knows that the godfather mentioned by the other three is not the same as the godfather in the mafia movie.

Rather, it refers to a person who, in a religious context, is given a religious name when he is baptized and undertakes to undertake his religious education.

Probably means that although the two Berkeley professors gave life to the DE algorithm, Chang Haonan and Xu Yang are the ones who gave the soul of the DE algorithm.

Considering that very few Chinese people can speak in this way, there should still be a certain degree of credibility in what the other party says.

For Chang Haonan, it is also a good thing to be named in these fields of engineering theories or algorithms that have nothing to do with classified projects, which is equivalent to putting on an excellent camouflage for him.

Truth and falsehood are the mystery of secrecy.

Otherwise, if you are a person who has done nothing on the surface, your status will jump up like a rocket, and it will not be difficult for people with a little brain to understand the reason.

At this time, the files in the floppy disk were also read.

Chang Haonan opened one of them.

"This is the first place in last year's ICEO competition, Princeton University's algorithm, and the following file is the second place, Kyoto University, because Nagoya University did not participate, so they are the hosts last year, and these two schools will also participate in the second edition this year, so they should be our main competitors."

Lin Zhi introduced in a timely manner beside him.

"1st place...... I'll take a look. ”

Chang Haonan opened the first program file inside.

"These two algorithms are indeed very high in optimization efficiency, we used the optimized DE algorithm last month to improve the efficiency of these five test functions, although compared with last year's third place has been greatly improved, but there is still no way to compete with them, so I thought I would like you to take a look, is there any further way."

Kan Feiang said and picked out another one from the floppy disk box:

"That's what we did."

"Don't worry, take a look at them first, they can be so much faster than DE, and they are quite capable."

In fact, when he heard that the differential evolution algorithm only won the third place just now, Chang Haonan was still a little surprised.

Even before he made improvements and optimizations, the efficiency of DE should in most cases surpass the more popular analog annealing algorithms of the era and the genetic algorithms that were about to become popular in the original timeline.

Considering that the test functions just looked at are all normal, there is nothing strange about it.

Only taking the third place is either because the contestant is not good at learning, or there is some other reason -

If there really was such an awesome algorithm, it would have been famous in the academic world a long time ago.

Because the time for the game is limited, and the test function is not complicated, it didn't take Chang Haonan long to watch it.

Of course, normally after reading it, the key parts will be analyzed and tested, but with his current theoretical level close to LV3, he can save the effort of the previous step to a large extent.

"This thing ......"

Chang Haonan frowned, and first tried to run the Michalewiczs function in the test question.

It is a relatively uncommon type of benchmark function, at least in this era, because it is a single-objective optimization function with several local minima-values and planar regions, where all z-values are nearly equal, which is not friendly to most algorithms.

The results are really good.

"We've tried several ways, but they can't run so efficiently, you can run the fifth test function again, and the results are even more exaggerated, so there have always been suspicions in the past year that these two groups cheated in the test, and about last week, the organizers were pressured to release all the relevant documents."

Song Tingyu, who had never spoken just now, said in a somewhat hoarse voice.

This explains why they insisted on DE even when they knew there was a more efficient algorithm.

It's not because they're admirers of the DE algorithm, but because they've always thought someone else was hanging up.

In any case, they made the right choice.

Because Chang Haonan has already understood the little tricks played by the other party.

Compared with the suspected opening and the real opening, it is still a little less interesting after all.

"No, I probably understand how they did it."

As he spoke, he opened the second-ranked file and skimmed through it faster than he had just done.

"Huh?"

Lin Zhi's surprised voice was just halfway out, and he quickly covered his mouth.

They originally planned to come over today to hand over the things to Chang Haonan, and then go back and wait for a few days of news.

Just now, I was ready to take out the gifts I brought and leave.

The result turned out to be on the spot?

"As long as we change the way we test them, these two algorithms will be completely revealed."

Chang Haonan moved his body and began to build a new mathematical model.

Min f(r，n)=-∏(i=1，m)R(ni)

. g1(r，n)，g2(r，n)，g3(r，n)

0≤ri≤1，ni∈Z+，1≤i≤m……

Despite having no background and only a few broad mathematical formulas, the trio were worthy of being selected by Yan to participate in international competitions, and Kan Feiang was quick to see the practical problems corresponding to this model:

"This should be a reliability redundancy calculation for a tandem system?"

Chang Haonan hasn't even written the annotations of the physical quantities represented by the symbols, so that it can still be seen, which on the one hand shows that his brain is really turning fast, and on the other hand, it also shows that there should be a lot of brushing questions in the past.

Physics is not pure mathematics, and no matter how smart a person is, it is impossible to see through physics problems that he has not been exposed to before.

"That's right, but don't underestimate this simple tandem system."

Now that they have understood, Chang Haonan is also happy to save the trouble of explaining, and after running the first algorithm to start optimization, he nodded and replied:

"There are many complex equipment, such as gas turbines or internal combustion engine overprotection systems, corresponding to this model, how to squeeze the potential of the equipment as much as possible without affecting the safety of the system, and reduce the redundancy has become the optimization goal of engineers."

"In the past, this thing can only be done by experience, so giants such as Siemens, Mitsubishi Heavy Industries, General Electric, and Ansaldo have entered the industry early, accumulated more data and experience, and even with hardware equipment of the same technical level, their unit performance indicators are much better, which brings them more customers, and in turn provides more feedback data, forming a virtuous circle, so in this model, later people can never surpass them."

"However, with the efforts of our algorithm engineers, this kind of control problem has gradually begun to move away from the pure empirical category in recent years, and we can get some theoretical guidance, so that the development efficiency of the control system can be accelerated, which also gives us those countries that have not caught up with the industrial revolution the possibility of catching up."

In fact, what Chang Haonan said is exactly what he has been doing in the past year-

Not only through his forward-looking vision, but also through technical means to improve the efficiency of scientific research, so that the limited funds can play a greater role than others.

After hearing this, the three people next to him were stunned at the same time, and then showed different, but obviously thinking expressions.

At the end of the 90s, the domestic computer major was also booming for a few years, but the foundation was not firmly shaken, and the heat of the computer science major during this period was quite blind, and the real first-mover dividend period in the early 90s has long passed, and the computer industry has not really developed.

As a result, these batches of students graduated just in time for the collapse of the Internet bubble, and encountered a group of people who were almost similar to the people who listened to the nonsense of "the 21st century is the century of biology" at the beginning of the 21st century.

Except for a few bigwigs, most people can't survive the industrial explosion more than ten years later, and they have already changed careers.

In this context, even undergraduates of a school of this level may not know what they are learning for, many people are simply accustomed to studying hard, or they don't need to study hard at all to get a high score, and then choose the major with the highest score.

After all, even in the information age more than 20 years later, the level of understanding of college entrance examination candidates about university majors is still low enough to make college entrance examination registration and career planning an independent industry, let alone now.

The three of them were actually only sophomores and had not seriously considered what they had learned before that.

Now listening to Chang Haonan say this, I really have some feelings for the first time.

As a few people spoke, the optimization algorithm running on the computer came to an end.

But unlike the previous ones, it's not really exporting anything.

"There's no need to go on."

Chang Haonan aborted the program and opened the log file, and quickly found the crux of the problem:

"There is a problem with the way to deal with constraints, so there is no way to converge, and no matter how much you run, you will not output a valid solution."

Three-faced .jpg in shock

"You see, although these two functions can't be counted as cheating, but in fact, they are also exploiting the loopholes of the competition system, the algorithm of the first place is only suitable for the competition verification function that can be quantified, and it cannot be called a general optimization algorithm at all, and it almost does not have the ability to solve complex practical problems, and the second place is slightly better, but to calculate the Latin side, it is not suitable for the case of too many parameters, and there are also great restrictions, in other words, the first two are test-taking algorithms born entirely for the competition."

Chang Haonan analyzed:

"Of course, from the perspective of the competition, general algorithms are not more noble than them, but personally, I still hope that the development of algorithms can be closer to the practical application level, and ......."

At this point, he hesitated, because what followed was indeed entirely his reasoning, but in the end he said it:

"And based on your descriptions, I guess this competition may add some restrictions to this behavior, such as adding a model or function outside the standard to the testing process."

“……”

In short, that night, Chang Haonan became the last person to leave the laboratory.

(End of chapter)