A Comparative Study of Clustering Algorithms for Profiling Researchers in Universities Through Google Scholar

This study compares the clustering algorithms used by four outstanding Iraqi institutions in Anbar, including Anbar University, Fallujah University, AI-Marf University and AI-Huda University. Meta-studies for clustering, studies by themselves ensure clustering, clustering Reinforcement learning, definition including clustering, and deep clustering include methods to be considered The analysis is based on three basic metrics of cluster performance: the Silhouette score, the Calinski-Harabasz index, and the Davies-Bouldin index. The collected evaluations show unique performance characteristics for each algorithm, which provides helpful information for algorithm selection in an educational setting. Meta-getting to know for clustering is an effective method. It works well on all parameters with a silhouette rating of 0.4655, a Calinski-Harabasz index of 189.8185, and a Davies-Bouldin index of 0.8202, just like 0.7287. The locating, which is supported by way of silhouette rankings and equal Calinski-Harabasz and Davies-Bouldin indexes, suggests that the efficacy of all 3 of those methods compares nicely inside the context of experimental measurements from the s. Deep clustering, alternatively, plays in another way, with an extensively decreased silhouette score of 0.0880 and extensively higher Calinski-Harabasz (83.4026) and Davies-Bouldin (3.2493) indices. This shows that its clustering performance may be more severe when compared to the other strategies studied. The most successful method is meta-gaining knowledge of for clustering, which plays nicely on all parameters with a silhouette rating of 0.4655, a Calinski-Harabasz index of 189.8185, and a Davis- Bouldin index of 0.8202, just like 0.7287 Finding this, similarly determined by silhouette scores and almost same Calinski-Harabasz, and Davis- Bouldin indices, shows that the effectiveness of these three technology is comparable in the context of experimental measures of the. On the other hand, deep clustering has unique overall performance profiles, with a substantially lower silhouette score of 0.0880 and considerably higher Calinski-Harabasz (83.4026) and Davis-Bouldin (3.2493) indices This suggests that its clustering overall performance can be lower compared with the alternative techniques considered. The outcomes of this have a look at laying the muse for making knowledgeable choices on algorithm choice for facts evaluation in instructional institutions. The assessment's ranking gives a sensible path, with meta-studying for clustering taking the lead, accompanied by using a triad of self-supervised mastering, reinforcement mastering, explainable clustering, and in the end, deep clustering. These findings strengthen the information on algorithmic programs in educational environments and feature implications for universities in the Anbar Governorate.