FXCoral

Neural Networks in the prediction of catastrophes

portada230421.jpg Free Use Resource from Pixabay, edited with PowerPoint By @tocho2.

Hello friends of Hope ..!

Whenever I find something interesting for my professional development I feel like sharing it with you, and in this post I bring you something that I found fascinating. The first time I heard it, I simply did not believe it, I lean towards the exact sciences and some things sometimes seem like a simple utopia to me, I tell you about the prediction of natural catastrophes through the use of Neural Networks.

By the time I was finishing my engineering studies, it was said that earthquakes could not be predicted, however, other phenomena such as floods, torrential rains, storms and tsunamis if possible predict and I have verified it based on my knowledge and experience in the area. A few years have passed and I still claim that earthquakes cannot be predicted.

An earthquake is the response to a sudden action that occurs inside the earth's crust, in other words it is the result "in the form of a violent shaking" when the release of energy (pressures) in the earth's crust, also known as tectonic plate movement and fault, realistically how could we know when and how they will happen?

However, science and engineering (my favorite area) have evolved enough to consolidate enough records of "almost all" natural phenomena: earthquakes, tsunamis, floods, etc. We can also currently find a large amount of sophisticated equipment such as state-of-the-art sensors and receivers that can find alterations in weather behavior patterns, for example, and that can alert scientists and observers of these phenomena about a possible occurrence of an extreme event.

clima230421.jpg Free Use Resource from Pixabay.

The same does not happen with earthquakes because their behavior pattern is not cyclical, as is the case of river flooding and the floods that this causes, the rains extreme snowfalls, these cases have return periods, determined duration and frequency of occurrence that have already been measured and recorded in previous years and when they occur again they do so with the same pattern of behavior, so they can easily be predicted . In contrast, earthquakes are far from having a known pattern of behavior.

Now I find that through the use of neural networks it could be possible to predict at least the small replicas and this would somehow contribute to fill the information gaps that still exist about these phenomena , as well as the behavior of all natural phenomena that claim lives and generate unfortunate catastrophes every time they happen unexpectedly could be better understood.

separadortocho2.jpg

What kind of contribution can Neural Networks provide in the prediction of catastrophes?

red230421.jpg Free Use Resource from Pixabay.

Taking into account that neural networks are an artificial replica of the nervous system of humans and animals, and remembering that this interconnected neural network feeds on information and data to emit a “ output "or response, as software does, we can say then that a neural network fed with information and fundamental records about some natural phenomenon, it is very likely that, for example, a response will be obtained with a possible prediction and simulation of the phenomenon.

The key is in the recognition of behavior patterns based on the exhaustive analyzes that the Neural Network would carry out in the databases and records with which it is fed, that means that It must be able to recognize periods of occurrence, magnitude and intensity of the phenomenon, durations, among other variables, making use of mathematical models from the simplest such as regression models that allow projecting and extrapolation to complex algorithms.

This means that the key is in the input data and all the information recorded from previous events ..!

I find all this fascinating because it constitutes a breakthrough on something that a few years ago was thought impossible or at least unattainable in the short term, fortunately I will be able to see it with my own eyes inside of little because the "future is now".

separadortocho2.jpg

In conclusion

I particularly think that the key to success is in the consolidation of extensive databases, I have seen this since before AI, Machine Learning and Neural Networks existed, since the records of previous events with all the variables that define each natural phenomenon and the catastrophes that they cause, have a relationship with each other that has not yet been deciphered 100%, it is here that scientists and engineers (I include myself here ) we rely on the use of Neural Networks and AI, at least we can take positive advantage of this. For the benefit and safety of our cities.

If you liked this post, comment and let me know your opinion ..!

More information on this topic here

baner 2.jpg

banner 1.jpg

GIF_MJ_3.1.gif

https://www.kastorianiestia.gr/2018/07/την-2η-θέση-στον-κόσμο-κατέκτησε-η-ομάδα/

Technology

Future reading