Temperature prediction system

1. A temperature prediction system comprising:

a data input module configured to receive data related to climate;

a prediction module having installed therein a trained model for predicting a temperature based on input data from the data input module; and

an output module configured to output temperature information predicted by the prediction module,

wherein the input data comprises at least one of temperature, humidity, wind speed, wind direction or precipitation,

wherein the trained model is trained using both observed data and a regional data assimilation prediction system (RDAPS) model,

wherein the trained model is provided by an artificial intelligence apparatus, the artificial intelligence apparatus comprising:

a first feature value extractor configured to extract a first feature value from the observed data, the first feature value being a temporal feature value;

a second feature value extractor configured to extract a second feature value from the RDAPS model, the second feature value being a spatial feature value; and

a merger configured to merge the first feature value and the second feature value, the merger providing the trained model.

2. The temperature prediction system of claim 1 , wherein the observed data is learned by Bi-long short terms memory (Bi-LSTM) networks.

3. The temperature prediction system of claim 1 , wherein the RDAPS model is learned by a convolution neural network (CNN).

4. The temperature prediction system of claim 3 , wherein the RDAPS model is learned using temperature image information.

5. The temperature prediction system of claim 1 , wherein the merger merges a vector of the first feature value and a vector of the second feature value.

6. The temperature prediction system of claim 1 , wherein the second feature value is merged by the merger, by extracting data similar to the temperature information using temperature information of the observed data as a key.

7. The temperature prediction system of claim 1 , wherein the input data comprises at least temperature information.

8. The temperature prediction system of claim 7 , wherein the input data comprises at least two of temperature, humidity, wind speed, wind direction and precipitation.

9. The temperature prediction system of claim 7 , wherein the input data comprises all temperature, humidity, wind speed, wind direction and precipitation.

10. A temperature prediction system comprising:

a data input module configured to receive data related to climate;

a prediction module having installed therein a trained model for predicting a temperature based on input data from the data input module; and

an output module configured to output temperature information predicted by the prediction module,

wherein the input data comprises at least temperature information,

wherein the trained model is trained using both observed data of a plurality of observation stations and a regional data assimilation prediction system (RDAPS) model,

wherein the RDAPS model is learned using information representing a temperature,

wherein the trained model is provided by an artificial intelligence apparatus, the artificial intelligence apparatus comprising:

a first feature value extractor configured to extract a first feature value from the observed data, the first feature value being a temporal feature value;

a second feature value extractor configured to extract a second feature value from the RDAPS model, the second feature value being a spatial feature value; and

a merger configured to merge the first feature value and the second feature value, the merger providing the trained model.

11. The temperature prediction system of claim 10 , wherein the observed data is learned by Bi-long short terms memory (Bi-LSTM) networks.

12. The temperature prediction system of claim 10 , wherein the RDAPS model is learned by a convolution neural network (CNN).

13. The temperature prediction system of claim 10 , wherein the RDAPS model is learned using temperature image information.