Part 2/7:
The research, spearheaded by astronomer Jean Dav from the German Aerospace Center, harnessed simulated data generated by a sophisticated system from the University of Bern, Switzerland. This system constructed thousands of artificial planetary scenarios to train the algorithm in recognizing patterns indicative of potentially habitable worlds.
Impressive Findings
The results, published in the journal Astronomy and Astrophysics, are remarkable. After being trained on over 53,000 simulated systems, the model has achieved an astounding accuracy of 99% in predicting which stars are likely to contain Earth-like planets. Key indicators analyzed included the mass, radius, and orbital period of the closest planet to each star.