Table 5 Comparative analysis based on various criteria
From: A brief survey on big data: technologies, terminologies and data-intensive applications
Ref | Technique | Advantage | Disadvantage | Achievements | Tools | Application |
|---|---|---|---|---|---|---|
[95] | Selection of optimal features for the applications of big data | It is efficient in the reduction of obstacles present in the feature selection model | The rate of convergence is low | Compared to other utilized algorithms' the accuracy levels of PSO and GWO are 86.8 and 81.6, hence obtained the average accuracy is 90 percent | MATLAB | Data mining |
[46] | Analysis based on middleware to overcome the performance issues using machine learning | In controlled environments, establish realistic application workloads | The index will use more memory and slow write operations to maintain the secondary index | Data-intensive application | ||
[96] | To provide better quantity service for intensive data applications in a cloud environment using Aneka | Minimum usage of resources while computing the risk task | Cost of scheduling is high | Basic Local. Alignment Search Tool (BLAST) | data-intensive application | |
[75] | Coreset-based data prioritizing solution to address security challenges brought on by jamming attacks | Improved quality of cluster with high detection efficiency | Inability to recover from database corruption | Based on the analysis of the quality index, the method obtained 0.805 using the Davies-Bouldin index (DBI) | NS-2 | VANET |
[98] | The use of the Non-dominated Sorting Differential Evolution (NSDE) algorithm—to increase the general superiority of the placement methods | The method achieves better load balancing and power consumption | Reduced latency for storage systems | MacBook Pro 2019 | IoT-based power electronic applications | |
[99] | Accelerating computing and data mining operations in the cloud | Provides highly efficient approximate computing | Fails to contribute to reaching the highest speed and complexity | Hadoop speedup of 17 × and Sappox speedup of 8 × with 5% error tolerance | CUDA | Applications of Data mining |
[23] | The performance of data-intensive applications using unstructured data using the Spark framework for hybrid program analysis | Improved performance on optimized code | Low performance due to the maximum size of data | The application can be accelerated by 7.47% and 2.967% more quickly with EP and CM | Data-intensive application | |
[90] | Workload allocation for data-intensive applications using the Energy Efficient and Bandwidth-Aware (EEBA) | Improved QoS and workload | High computation time | High cloud QoS with 11%, 38%, and 15% rate of makespan augmentation using Simple, Mixed, and Heavy BoT | CloudSim | application with a high data volume |
[100] | Azure cloud analysis of physics data intensively | Increased efficiency and quality of research | High collection of data utilized with high computing time | For CPU doubling the CMS UCSD resources deliver a 7 M core, and the cloud-enabled CE provides 7.3 M hours each month | Grid Community Toolkit (GCT) | Data-intensive application |
[30] | Spark Streaming Analysis for Data-Intensive Pipelines | Provides high throughput with quick efficiency of data transfer | High memory consumption | SUM Statistics for the GC The throughput of the Server Application with Backpressure is 100% | GC Analyser | Data-intensive application |
[37] | A cross-point array dubbed with XAM serves as the foundation of Monarch, a resistive 3D stacked memory | Cost efficiency with improved performance | High latency | In String-Match, the better performance of Monarch occurs 24 × , 11 × , 12 × , and 14 × in HBM-SP, CMOS, HBM-C, and RRAM | ESESC multicore simulator | memory-intensive applications |
[88] | To bridge the gap between programming models, HPC languages, and Big Data using the IgnisHPC5 framework | Improved performance with efficient execution | Lack of resources | MSAProbs produce a maximum performance difference of 0.4 percent | Python | MPI based applications |