This article analyzes some issues that can be innovated
Refer to other people's articles, basically just use machine learning to do things that others have not done, or the effect will be better.
list some pain points
1. For the use of intelligent marine traffic, the accuracy must be ensured
2. It needs to be real-time, and the next position can be predicted immediately after the ship's position returns.
3. The model needs to evolve, what if the ship transforms or travels occasionally
4. The machine learning model requires a large amount of data training, and the cold start problem must be solved.
5. The data preprocessing method is mature, but the solution to the problem is innovative.
Supplement: Causal g of independent variable and dependent variable
Now it is becoming more and more clear that data has three main characteristics: nonlinearity, big data, and streaming data
Innovations can be new processing architectures, new predictive models, new data preprocessing methods