Scientists use artificial intelligence to predict the possibility of life on other planets

According to a recent study by a team at the University of Plymouth, the development of artificial intelligence can help scientists predict the possibility of life on other planets. This study uses artificial neural networks (ANNs) to classify planets into five categories, estimating the probability of life in each case. Can be used for future interplanetary exploration missions. This work was published on April 4th by Christopher Bishop at the European Astronomy and Space Science Week (EWASS) in Liverpool. The artificial neural network is an attempt to replicate the way human brain learns. System. They are one of the main tools used in machine learning and are particularly good at identifying complex patterns that are very complex for the biological brain.

This composite image shows an infrared image of Saturn's satellite Titan, taken from NASA's Cassini spacecraft. Some measures have shown that Titan has the highest inhabitable rating in any world except Earth, which is based on energy. Availability, and various surface and atmospheric characteristics. Image Credit: NASA / JPL / University of Arizona / University of Idaho

The team is located at the center of the robot and nervous system of the University of Plymouth. They have already 'trained' this network. According to whether it is most like the Earth, early Earth, Mars, Venus or Saturn’s satellite Titan, the planet is divided into five different types. Types. These five are all rocky bodies known to have the atmosphere, and are one of the most habitable planets in the solar system. Mr. Bishop commented: We are interested in these ANNs now because they are on an imaginary smart star. The spacecraft explored, this is a scan of the extrasolar planet system.

Study large-area, deployable, planar Fresnel antennas for use in returning from distant interplanetary space probes to Earth. If this technology were to be used in robotic spacecraft in the future, this would be necessary. Atmospheric observations - known as The five solar system objects of the spectrum were taken as input to the network and were then asked to classify them according to the type of planet. Since the known life exists only on the earth, the classification uses the 'life probability' metric, which is based on Relatively good atmospheric and orbital properties for the five types of targets. Bishop exploits hundreds of different spectral lines to train the network. Each spectral line has several hundred parameters. These parameters help to adapt to living. surroundings.

These inputs represent the values ​​from the spectrum of the test planet's atmosphere. The output layer contains a 'probability of life', which is based on the measurement of the input's similarity to the five solar system targets. The input passes through a series of hidden layers in the network. These layers are Interconnected, enabling the network to 'learn' which patterns of lines correspond to particular types of planets. Image Credit: Bishop / Plymouth University

So far, the network has performed well when presenting spectral profiles of tests that have never been seen before. Dr. Angelo Angelozi, the project's director, said: Taking into account the current results, this method may prove to be Classifying different types of exoplanets is very useful. They use results from terrestrial and near-Earth stations. This technique may also be suitable for selecting targets for future observations, taking into account upcoming space missions such as the ESA Ariel space mission. And NASA's James Webb Space Telescope Spectral Details Increased.