Early Prediction of Cyber Hacking Breaches in Network Using the Novel CNN GoogleNet Compared Against AlexNet Machine Learning Algorithms

The aim of this research study is to predict the Cyber Hacking Breaches at earlier stage using dataset. For this two machine learning algorithms Novel Convolutional Neural Network (CNN) GoogleNet and AlexNet algorithms were used. Novel CNN GoogleNet and AlexNet are using training and testing dataset splits of cyber hacking breaches. There are 2030 cyber hacking breach images collected for this research study from an open source kaggle web resource. The dataset was split the data was divided into two sets, with 80% of training dataset and a 20% testing dataset. The train data set consists of 1624 images and test data consists of 406 data set images. The g-power test used is about 80% g-power setting parameters:
$\alpha = 0.05$
and power=0.85. The experiment study was iterated twenty times using the above said models. This research study has found the Novel CNN GoogleNet algorithm has a 93.40% increased the accuracy, when contrasted with the AlexNet algorithm value 80.76%. The statistical significance level for the T-Test for Independent Samples (p=0.000, p<0.05) with a 95% assurance interval. The CNN GoogleNet has better significant value than AlexNet. The outcome shows has the earlier prediction of Cyber Hacking Breaches. When Comparing the Accuracy of results, it is showing significant improvement on CNN GoogleNet compared over AlexNet in predicting cyber hacking breaches in the early stage.