Diseases have been prominent cause of deaths. Early detection of diseases can exponentially decrease the death rates. Among many diseases, Lung cancer needs early stage identification to save life of human being. If it is not identified in the early stage then it causes the death of the human. The treatments during early stages have proven to be effective and saved the life of humans. Now the problem is to identify the lung cancer disease at early stage. Deep Learning plays a major to find the early stage of this disease from the given Computed Tomography (CT) images accurately. It is one of the popular and effective approaches in classification problems because it uses transfer learning and convolutional neural network. The accuracy for these kind problems can be increased by proper tuning the necessary parameters of model. This work focuses on implementation of pre-trained convolution neural network models which include Alexnet, VGG, ResNet-50, Densenet-121, Inception, and MobileNet with proper fine-tuning of the necessary parameters. In this work, the benchmark dataset "IQOTHNCCE lung cancer CT scans images" is considered to check the performance of these pre-trained models. Each of the implemented models is evaluated on the basis of performance metrics such as Balanced Accuracy Score, Precision, Recall, and F1 score. The Inception model performed best with balanced accuracy score of 99.8% among other models.