Browsing by Author "HAIDER Ali"

Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Examining deep learning and machine learning models for crime prediction
    (UMT Lahore, 2022) Ahmed Tariq; HAIDER Ali
    The world is progressing day by day. Technology is improving and doing wonders, however, this world has many problems that are still unsolved. Crime is one of those big problems. Crime is increasing and it is getting out of hand in many countries. Criminal activities and violence is affecting the lives of people, therefore, disturbing the country’s peace. With the advancement of technology, some measures are taken to control crime. Law enforcement authorities are using the latest technologies to study crime data and predict crime. Machine learning and Deep learning are those technologies that help them examine crime records and forecast crime. In this project, we have used these two technologies to examine crime data and predict crime. This study applied different Machine-Learning algorithms like XG Boost, Random Forest, KNN, Ada Boost, Decision Tree, etc. These models are used to predict crime on the Chicago Crime dataset. The performance of the Random forest outperformed from all the models. The experimental results exhibited that the presented model is feasible and efficient. The execution of per algorithm is estimated using the metrics such as precision, recall, and F-measure, and the results have resembled. Moreover, we reckon the efficacy of other parameters in deep learning architectures such as NN, DNN, and CNN. Deep learning gives better results but the CNN model gives better accuracy. We have figured out that the boosting models can perform well on large datasets but up to some extent whereas Deep learning models can perform betteron a large dataset and take less time