Computer Science and Genetic Algorithms

We all know about the importance of technology and Artificial Intelligence in today’s society. However, some people think that Artificial Intelligence is a very difficult science and requires a lot of knowledge. Nowadays, Artificial Intelligence has many applications and approaches and it can solve multiple problems. For example: Genetic Algorithms. The evolution property of life forms in optimising problems inspire these algorithms. In other words, it is a set of procedures that try to figure out the most optimal solution from millions of solutions. Do you want to know more about this problem-solving approach in computer science? Keep reading!

Natural Selection

How can nature create insects that are resistant to pesticides or snakes with skin colors that adapt to the place they live in? This is a process of generations and generations of these species evolving through time. Living things that survive nowadays are considered the optimal state in their present environment. That is because their ancestor survived many challenges in the past. Living things are constantly evolving.

For example, we have got antibiotics that used to work on bird flu, but they no long do. This means that viruses evolve quickly and become drug resistant. This is the result of natural selection. The most powerful viruses survive and the rest are eliminated. Because of this, the surviving viruses will become hybrids and create their next generation with their genes. That next generation will be immune to antibiotics. It was also noticed that there are 3 important characteristics of each generation of viruses. Selection, crossover and mutation.

Imitating Natural Selection in Computers

This natural selection process is used by computer scientists in order to find the most optimal solution out of many options. But, how does it work? They select at random multiple solutions on computers. After that, they assess their performance in each run (generation) and eliminate the weaker solutions in this generation. Then, the scientists combine the features of the surviving solution in each run and make random small mutations to it. They use this modified solution for the next run. Computer scientists usually run these assessments many times for the purpose of getting an optimal performing solution from all generations.

The best solutions will result automatically from the implementation of natural selection. This process does not require any human selection. Because of this, we can classify this process as Artificial Intelligence. Thanks to the improvements of computers and technology, there exist many applications of AI. However, Genetic evolution Algorithm is very useful and accessible nowadays.

Can we apply it to real-life problems?

Genetic Algorithms can solve many problems. However, we have to provide enough computer power in order for this to happen. There are many applications of Genetic Algorithm (GA) nowadays.  Right now, we will show you 3 examples:

• Design Simulation Application

Genetic Algorithm can design automatically an antenna used as spacecraft. In order to do so, scientists have ran many simulation seeking the best performing shape of  the radio antenna. We can see it as a wind turbine even more effective than the one that humans built. However, this antenna it is just design for experimental purposes. The simulations done, were computer based only.

• Graph Application

There is a famous graph mining problem called Travelling Salesman Problem (TSP). A salesperson has to travel to all the cities (red dots). However, there are billions of possible combinations that he can choose to travel to all points. Some are very long and others are shorter. What is the solution? Genetic algorithm can discover a reasonably short path for the salesperson.

• Artificial Intelligence Gaming

Genetic Algorithm can also have application in multiple video games. For example, it can train the best Flappy bird Artificial Intelligence player. You can see how it works in this video. Artificial Intelligence ran through multiple generation to dominate the game.

What does the future holds?

There are still lots of programs and mechanism that need to be discover by humans. Genetic algorithm is just one of the many life science methods that wants to solve modern problems. Furthermore, Artificial Intelligence is a technology that is developing more and more everyday. That is why is necessary to have scientists and experts for the future challenges that will arise.

Are you interested in Artificial Intelligence? Do you want to learn everything about in order to work as an AI scientist in a few years? This is your moment! In the Universidad de Alcalá we give you the chance to enrol in the Master in Artificial Intelligence we offer. Along with the greatest experts you will develop into the professional you wish to be. Do not hesitate and ask us for any information you may need. We will wait for you in Alcalá!