Chapter 370 (1/2)

Chapter 373

it took more than ten minutes for Cheng Nuo to read the document sent by Professor Fresnel.

The requirements of MI Xing company are very simple, at least Cheng Nuo looks like this.

They only need a set of procedures that can design a set of company financial and stock income, as well as individual stock investment, according to the specific situation of their company.

This is the most basic data program for all financial securities companies.

There are a lot of information about this on the Internet. Cheng Nuo has a general look at it.

He leaned against his chin, frowned, and thought for a while, and was immediately inspired.

Cheng Nuo has made his own decision about the algorithm used to run the program.

Ga-rs-lr algorithm!

The language translated into Chinese is genetic algorithm rough set logic regression method.

This is a blind deletion attribute reduction algorithm and heuristic algorithm, which was first proposed by Polish mathematicians in 1982.

When dealing with vague and uncertain information, the mathematical tool has a certain effect.

For dealing with the fuzzy data relationship between corporate finance and stock returns in this project, this algorithm is no longer suitable.

Time is tight, only seven days.

Now that the core algorithm is determined, the following is the program modeling plan.

Cheng Nuo thought about the project for a while, clearing up the idea of the project: using ga-rs method to obtain the optimal reduction of financial indicators, and then using LR model to explore the relationship between the two. Finally, through the reduction of ga-rs, we get the impact of financial indicators on stock investment.

At this point, the main part of the project should have been completed.

Cheng Nuo a corner of the mouth, a new document, fingers on the keyboard, began to work.

This kind of difficult modeling project is not qualified to let Cheng Nuo use the draft paper to calculate directly in his mind.

Cheng Nuo clattered on the keyboard, thinking as smooth as gurgling water.

Genetic algorithm can be global optimization and implicit parallel, but there are still specific factors to be considered.

The first problem is chromosome coding.

Let the set of conditional attributes be C = {C1, C2, C3 Then the conditional attribute space Ω C can be mapped to the chromosome of genetic algorithm. Each chromosome is represented by n-dimensional (0,1) binary bit string, which makes the chromosome individual and attribute subset establish a one-to-one correspondence relationship.

Then the fitness function is constructed, which can be combined with relative dependence to satisfy the principle of attribute reduction:

F (x) = f (x) γ CD (x) = (1-card (x) n) * card (POS (d)) card (U).

the remaining factors include selection probability, crossover and mutation, termination condition of algorithm, etc.

But Cheng Nuo has not yet done this part of the work, was handed out by hull to lunch, along with Henry and gal.

Along the way, Cheng Nuo and hull walked in front of each other and chatted. Henry and gal kept staring at Cheng Nuo's back, as if to see something from Cheng Nuo's back.

Unfortunately, they didn't see anything.

They often see Chinese people in MIT.

But in the eyes of two people, Cheng Nuo is more handsome and younger than those people. There is no difference between them!

Why did he take over a project that the three of them couldn't finish in a week?

Both of them have been thinking about this question all morning, but they haven't figured it out.

Talking and laughing, Cheng Nuo and hull walked into the lunch place School canteen.

Yes, it's the school canteen.