Modified DMGC Algorithm using ReLU-6 with Improved Learning Rate for Complex Cluster Associations

This research work has presented a modified deep multi-graph clustering (MDMGC) for the enhancement of clustering accuracy. Multi-graph clustering is an important method for shallow models. Shallow models are known to have highly complex data. The proposed MDMGC is based on a graph auto encoder with a deep neural network. ReLU-6 is used as an activation function in the proposed method with a learning rate of 0.0002. A loss function is used for performance improvement. In this proposed MDMGC. T-distributed stochastic neighbor embedding (TSNE) uses data set visualization as well as embedding. Use Google Co-Lab for the implementation of the proposed method. The proposed modified clustering method shows better results as compared to other methods in terms of different parameters, normalized mutual information (NMI), and loss function. There are different data sets available to perform the proposed work; they are 20 news and DBLP. The proposed method also shows better clustering visual outcomes when it is calculated with the help of TSNE.