Machine learning techniques based on security management in smart cities using robots

Nowadays, the growth of smart cities is enhanced gradually, which collects a lot of information and communication technologies that are used to maximize the quality of services. Even though the intelligent city concept provides a lot of valuable services, security management is still one of the major issues due to shared threats and activities. For overcoming the above problems, smart cities' security factors should be analyzed continuously to eliminate the unwanted activities that used to enhance the quality of the services.

To address the discussed problem, active machine learning techniques are used to predict the quality of services in the smart city manages security-related issues. In this work, a deep reinforcement learning concept is used to learn the features of smart cities; the learning concept understands the entire activities of the smart city. During this energetic city, information is gathered with the help of security robots called cobalt robots. The smart cities related to new incoming features are examined through the use of a modular neural network.

The system successfully predicts the unwanted activity in intelligent cities by dividing the collected data into a smaller subset, which reduces the complexity and improves the overall security management process. The efficiency of the system is evaluated using experimental analysis.

This exploratory study is conducted on the 200 obstacles are placed in the smart city, and the introduced DRL with MDNN approach attains maximum results on security maintains.