# Future Prediction Algorithm

## Recommended Posts

Would it be theoretically possible to construct an algorithm that when provided with sufficient data, could predict the future with reasonable accuracy based on trends and pattern prediction? For example, if you created an algorithm that predicts where and when an earthquake is likely to happen up to a year in advance, then provided it with all the available information on earthquakes and plate tectonics and so on.

##### Share on other sites

What your describing is physical models. So trivially, yes, depending on accuracy and ignoring your example of earthquakes.

To give an example slightly closer to earthquakes, then numerical weather forecasting?

##### Share on other sites

2 hours ago, StarKnight said:

Would it be theoretically possible to construct an algorithm that when provided with sufficient data, could predict the future with reasonable accuracy based on trends and pattern prediction? For example, if you created an algorithm that predicts where and when an earthquake is likely to happen up to a year in advance, then provided it with all the available information on earthquakes and plate tectonics and so on.

Probably not.  I suspect earthquakes like weather have a dependance on environmental variables that are inherently unstable.  I can tell you that the average temperature in Jan 2 is going to be colder than July 2, but it will probably never be possible to forecast the temperature or precipitation out more than about 14 days.

##### Share on other sites

On 6/16/2019 at 8:54 PM, Bufofrog said:

I ﻿can tell you that the average temperature in Jan 2 is going to be colder than July 2,

..not without knowing longitude, latitude, altitude etc. etc. and so on..

What is true on Northern Hemisphere is not true anymore on Southern Hemisphere.

##### Share on other sites

On 6/16/2019 at 11:02 AM, StarKnight said:

Would it be theoretically possible to construct an algorithm that when provided with sufficient data, could predict the future with reasonable accuracy based on trends and pattern prediction? For example, if you created an algorithm that predicts where and when an earthquake is likely to happen up to a year in advance, then provided it with all the available information on earthquakes and plate tectonics and so on.

Only I know that process

I suspect after Beethoven went deaf he might have taken Newton's work and came to the same conclusions. How else could his symphonies arouse emotions better than any other Classical Composer unless he could predict which notes would have the highest overall affect mathematically? After-all he would have had to of either remembered the notes before he went deaf, when he wasn't even a well-liked composer, or he was somehow able to know what notes would go together to cause an effect similar to how the Ancient Egyptians used a Nilometer to treat the rise and fall or the Nile during the flood season. This allowed them to predict which crops would be most successful in the coming harvest.

##### Share on other sites

3 hours ago, Sensei said:

..not without knowing longitude, latitude, altitude etc. etc. and so on..

What is true on Northern Hemisphere is not true anymore on Southern Hemisphere.

Thanks for clearing that up.

## Create an account

Register a new account