Configuration and Management of Machine Learning Service

Feedback

The machine learning service realizes AIGIS capability and provides a variety of algorithm support in the field of AI, such as ground object classification, target detection and so on. This section describes how to configure the machine learning service.

Note: The machine learning service does not support Web license at present. Please refer to the relevant steps to configure normal license or cloud license.

When deep learning functions such as binary classification, ground object classification, target detection, scene classification and object extraction are used, the following environment configuration is performed, and the machine learning functions related to clustering, classification and regression are not required to be configured.

  1.  Configuring the Python runtime environment

The machine learning service relies on the support of the Python runtime. iServer currently supports the following ways to configure the Python runtime environment:

  1.  Configuring machine learning resource bundles
  2. Download the machine learning resource package (supermap-iobjectspy-resources- ml. Zip) from the "Component GIS" list in the technical resource center of the official website, extract it to a resources-ml folder, and place it in 【SuperMap IServer product directory】/support/python. The resource kit provides sample data and sample models for reference.
  1. Install the system libraries

In the Linux system, in addition to the above configuration, you also need to install the libsm.so.6 system library, and following is an example of installing libsm.so.6 on Ubuntu.

If the computer can be connected to the Internet, directly execute the following command to install the libsm.So.6:

apt-get install libsm6

Please start with the networked environment https://pkgs.org download the installation package corresponding to the system and number of bits, and then proceed with the installation