Chapter 422 Professor Xu, I have some questions
Resource allocation can greatly improve the utilization efficiency of various resources such as energy.
If artificial intelligence can shine in these areas, human life will change greatly in the future.
The last item, competitive games, is relatively the fairest one and can reflect the computing power of artificial intelligence more objectively.
After these details were determined, the date of the competition was set for a month later.
"Professor Xu, isn't one month too short?"
Han Shubin knows that apart from the field of competitive games, we have not yet started research in the other two fields.
The technology of IBM and Google is relatively mature.
"One month is enough, let's set it at this time."
Seeing that Xu You was so confident, Han Shubin stopped asking questions.
In fact, even if the deadline is shortened to half a month, Xu You is very confident that he can complete the tasks.
Xu You can't wait for this quantum competition.
"Let's start with the weather forecast process."
Although Xu You had not formally conducted research in the field of weather forecasting using artificial intelligence before.
But Xu You is still very familiar with artificial intelligence's method of predicting weather.
In weather forecasting, the main methods used by artificial intelligence include intelligent grid forecasting and model analysis technology.
Smart grid forecasting uses big data analysis techniques to conduct all-round mining and analysis of large amounts of weather data.
Through such an intelligent grid system, artificial intelligence can accurately predict various weather conditions.
This weather forecasting method is accurate in predicting precipitation, but its disadvantage is that the forecasting period is relatively short and it is impossible to accurately predict the weather several days later.
Another way to predict weather is through model analysis technology.
Model analysis technology also uses various big data and artificial intelligence, but the focus is on the need to model a complex meteorological system.
This method of weather forecasting may not be as accurate as the first method in the short term, but it has a higher prediction accuracy in the long term.
The weather within a month can be predicted with relative accuracy.
However, because the amount of calculation is so huge, the performance requirements for the computer are very high.
Even many supercomputers are unable to handle such a huge amount of calculations.
But these are not a problem for quantum computers with extremely high computing speeds.
What's more, Xu You also has the skill of brain simulation to help him complete the modeling work.
In just three days, Xu You completed all the programming and modeling work and taught the algorithm to predict the weather.
"Our Suanjing weather forecast system has learned a number of weather forecasting methods, such as intelligent grid forecasting and model analysis technology, and can accurately predict the weather through its own system scoring mechanism. When there is sufficient meteorological data , Suanjing can predict the weather within 24 hours with almost 100% accuracy. Even for the weather within a month, the forecast accuracy can be increased to more than 95%."
Because there are so many factors that affect the weather, it is almost impossible to predict the weather a few days later with 100% accuracy.
A butterfly flapping its wings can change the weather on a certain day.
Not to mention, artificial rainfall and other human changes to the weather.
However, a researcher raised his own questions about the data provided by Xu You.
"Professor Xu, I can understand the accuracy of the weather forecast within 24 hours. But...how did you come up with the accuracy of the weather forecast within a month?"
This doubt is very normal because it has only been three days since Xu You made this weather forecast model.
There is no time to compile statistics on the accuracy of the model.
"This data is a theoretical value. We will know the specific accuracy later."
As he spoke, Xu You displayed on the big screen the weather forecast just made by the artificial intelligence algorithm.
Based on radar and other data provided by the National Meteorological Observatory, artificial intelligence has completed weather forecasts for all parts of the world within one month.
However, compared with the weather forecast given by the Meteorological Bureau, the weather forecast based on artificial intelligence will have some discrepancies. Even on a certain day in a certain place, whether it will be sunny or rainy may be given a completely different prediction.
"Professor Xu, if this is just a theoretical value, does this model lack sufficient verification?"
"We will observe for half a month first. If the data does not meet the standards, we will make changes to the model."
In fact, Xu You is very confident. According to the results of Xu You’s brain simulation, the accuracy of this model is even higher than the data given by Xu You.
Xu You also understood their doubts. After all, if the normal procedures were followed, multiple verifications and modifications would definitely be required.
"I agree with Professor Xu. We will know the prediction accuracy of the model in a few days," said Han Shubin.
Even Han Shubin couldn't understand how Xu You came up with the theoretical values predicted by the model.
But as long as this achievement comes from Xu You, there is nothing to doubt.
After completing the weather forecast model, Xu You then studied the task of resource allocation.
Compared with weather forecasting, resource allocation problems are much less random, and the main focus is on the computing power of quantum computers.
For example, in terms of energy allocation, the data provided by the power grid can be used to predict electricity load, and then predictive maintenance measures can be provided to provide accurate electricity supply and demand solutions.
Or in the field of wind power generation, based on historical power generation data and weather forecast information, we can build and train neural network models to optimize wind power generation plans and improve wind power generation efficiency.
Two days later, the AI had learned to solve various resource allocation problems.
Compared with previous models, computing artificial intelligence can improve efficiency by 20 to 50 percent, making resource allocation more reasonable.
As the past two days have passed, the accuracy of weather forecasts using artificial intelligence has also been verified.
"Professor Xu, the accuracy of our weather forecasts for various parts of the world in the past two days has reached 99.9%. Among the locations where the forecasts were inaccurate, there were many places where human actions such as artificial rainfall were carried out, which affected the accuracy of our forecasts," said a member of the project team.
Such accuracy means that artificial intelligence will only make one mistake in predicting the weather a thousand times.
This is already a very high number considering that weather forecasts are already full of randomness.
