Deep Learning for Big Data: A Survey of Architectures,Distributed Frameworks, and Emerging Challenges
TL;DR
This survey reviews the intersection of deep learning and big data along three axes: the neural architectures used to model large-scale data, the distributed computing frameworks that make training such models tractable, and the persistent challenges of scalability, data quality, privacy, andinterpretability that constrain real-world deployment.
Abstract
Big data has become a defining feature of moderncomputing, and deep learning has emerged as the primary toolfor extracting predictive value from such large-scale, high-velocity, heterogeneous data. This survey reviews theintersection of deep learning and big data along three axes: theneural architectures used to model large-scale data, thedistributed computing frameworks (notably those built onApache Spark) that make training such models tractable, and thepersistent challenges of scalability, data quality, privacy, andinterpretability that constrain real-world deployment. Wefurther examine emerging responses to these challenges,including federated learning for privacy-preserving distributedtraining, explainable AI (XAI) for large-scale models, andedge/TinyML approaches for resource-constrained deployment.The survey closes with open research directions, includingcommunication-efficient distributed training, robustness to non-IID data, and standardized benchmarking for big data deeplearning systems
