Contact me if you have a similar task.
Neural networks can establish connections between input and output through autonomous learning.
The challenge consists in skillfully preparing the input (feature engineering) and in setting up adequate network structures: architecture of the neural layers, ConvNet filters, activation functions.
A classic application is the recognition of handwritten digits:
Examples of use cases in industrial practice are the classification of objects on images, the detection of problematic financial transactions or the automatic topic assignment in large text corpora.
Contact me if you have a similar task.