Criminal Identification and Comprehensive Analysis Using Decision Tree Classifier

Machine learning is the process of deriving insights or new knowledge from large data sets. During a criminal investigation, cops collect a mountain of data, but only the most pertinent pieces need to be analysed. So, it’s possible to use Machine Learning for this function. The results produced are more sensitive to the choice of Machine Learning technique. This is the main justification for comparing and choosing the best performing Machine Learning algorithm. Anybody can grasp the fundamental idea of utilising a Decision Tree to categorise data. In addition, this paper uses Decision-Tree algorithms and face pattern analysis tools to study criminal faces. This was achieved in MATLAB by using face pattern tools that were embedded into the DT algorithms. The face detection algorithm in the MATALAB platform has also been upgraded to work in real time. Prior work has mostly focused on criminal event data sets but has made no adjustments for real-time functionality. The images of actual criminals have been used to train and test the face data sets that have been stored. In our experiments, the proportion of correct predictions of faces has increased by 2%. The test accuracy provided by the prior work is 78%, while the proposed method is 80%.