Random Forest Architectures on FPGA for Multiple Applications

Lin, Xiang; Blanton, R. D.; Thomas, Donald E.

doi:10.1145/3060403.3060416

Cited by 29 publications

(8 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Thanks to their inherent concurrent memory accesses and computational parallelism FPGAs are a good platform for performance boost of random forests. In [60] authors demonstrate the trade-o↵s between area utilization and context switch time between di↵erent architectures, and show how each architecture maps well to a di↵erent design scenario. The paper [61] compares and contrasts the e↵ectiveness of FPGAs, GP-GPUs, and multicore CPUs for accelerating classification using models generated by compact random forest machine learning classifiers and shows that FPGAs provide the highest performance solution, but require a multi-chip / multi-board system to execute even modest sized forests, while GP-GPUs o↵er a more flexible solution with reasonably high performance that scales with forest size.…”

Section: Related Workmentioning

confidence: 99%

Enhancing random forest classification with NLP in DAMEH: A system for DAta Management in eHealth Domain

et al. 2021

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

Enhancing random forest classification with NLP in DAMEH: A system for DAta Management in eHealth Domain

et al. 2021

View full text Add to dashboard Cite

“…Therefore, the highly parallel architecture of FPGAs is exploited to accelerate the design even further. In a different implementation domain, authors in [16] suggested two possible architectures for a random forest implementation on hardware, memory-centric and comparator-centric. The memory-centric approach enables a quick context switching from one random forest model to another through simply loading new node information into the tree level memory.…”

Section: Hardware Implementationmentioning

confidence: 99%

“…Keeping in mind the use of two datasets, this entails the need to incorporate the context switching feature of the memory-centric architecture. Therefore, in this paper, and with noticeable modifications to the implementation suggested in [16], the proposed implementation will follow the memory-centric approach.…”

Section: Hardware Implementationmentioning

confidence: 99%

FPGA-Based Network Traffic Classification Using Machine Learning

2020

View full text Add to dashboard Cite

Real-time classification of internet traffic is critical for the efficient management of networks. Classification approaches based on machine learning techniques have shown promising results with high levels of accuracy. In this paper, the suitability of packet-level and flow-level features is validated using stepwise regression and random forest feature selection. Moreover, the optimal percentage of packets considered within a flow while extracting flow-level features is determined. Several experiments are conducted using naïve Bayes, support vector machine, k-nearest neighbor, random forest, and artificial neural networks on the University of Brescia (UNIBS) and the University of New Brunswick (UNB) datasets, which are both publicly available. The performed experiments show that 60% of flow packets are a good compromise that ensures high performance in the least processing time. The results of the conducted experiments indicate that random forest outperforms other algorithms achieving a maximum accuracy of 98.5% and an F-score of 0.932. Further, and since software-based classifiers cannot meet the anticipated real-time requirements, we propose a Field-Programmable Gate Array (FPGA) based random forest implementation that utilizes a highly pipelined architecture to accelerate such a time-consuming task. The proposed design achieves an average throughput of 163.24 Gbps, exceeding throughputs of reported hardware-based classifiers that use comparable approaches, which in turn ensures the continuity of realtime traffic classification at congested data centers.

show abstract

“…In Ref. [17] three different architectures are proposed for a random forest classifier on a ZYNQ evaluation board, and in Ref.…”

Section: Related Workmentioning

confidence: 99%

FPGA and GPU-based acceleration of ML workloads on Amazon cloud - A case study using gradient boosted decision tree library

Shepovalov

Akella

2020

Integration

View full text Add to dashboard Cite

Random Forest Architectures on FPGA for Multiple Applications

Cited by 29 publications

References 11 publications

Enhancing random forest classification with NLP in DAMEH: A system for DAta Management in eHealth Domain

Enhancing random forest classification with NLP in DAMEH: A system for DAta Management in eHealth Domain

FPGA-Based Network Traffic Classification Using Machine Learning

FPGA and GPU-based acceleration of ML workloads on Amazon cloud - A case study using gradient boosted decision tree library

Contact Info

Product

Resources

About