A Deep Swarm-Optimized Model for Leveraging Industrial Data Analytics in Cognitive Manufacturing
Akshi Kumar, Arunima Jaiswal
Abstract
To compete in the current data-driven economy, it is essential that industrial manufacturers leverage real-time tangible information assets and embrace big data technologies. Data classification is one of the most proverbial analytical techniques within the cognitively capable manufacturing industries for finding the patterns in the structured and unstructured data at the plant, enterprise, and industry levels. This article presents a cognition-driven analytics model, CNN-WSADT, for the real-time data classification using three soft computing techniques, namely, deep learning [convolution neural network (CNN)], machine learning [decision tree (DT)], and swarm intelligence [wolf search algorithm (WSA)]. The proposed deep swarm-optimized classifier is a feature-boosted DT, which learns features using a deep convolution net and an optimal feature set built using a metaheuristic WSA. The performance of CNN-WSADT is studied on two benchmark datasets and the experimental results depict that the proposed cognition model outperforms the other considered algorithms in terms of the classification accuracy.