Online Extreme Learning Machine Design for the Application of Federated Learning

Chen, Yi-Ta; Chuang, Yu-Chuan; Wu, An-Yeu

doi:10.1109/aicas48895.2020.9073802

Cited by 12 publications

(6 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This increased learning without switching between domains may allow the network weights to travel further towards local optima for the industrial scale dataset in each epoch. This contrasts with results presented in the wider literature, where federated learning degraded model accuracy compared with non-federated learning by 3.3% [50], 1.66% [51], and <10% [52]. Figure 5e,f display the classification results for the previously discussed federated models fine-tuned on the industrial dataset.…”

Section: Machine Learningcontrasting

confidence: 63%

Domain Adaptation and Federated Learning for Ultrasonic Monitoring of Beer Fermentation

2021

View full text Add to dashboard Cite

Beer fermentation processes are traditionally monitored through sampling and off-line wort density measurements. In-line and on-line sensors would provide real-time data on the fermentation progress whilst minimising human involvement, enabling identification of lagging fermentations or prediction of ethanol production end points. Ultrasonic sensors have previously been used for in-line and on-line fermentation monitoring and are increasingly being combined with machine learning models to interpret the sensor measurements. However, fermentation processes typically last many days and so impose a significant time investment to collect data from a sufficient number of batches for machine learning model training. This expenditure of effort must be multiplied if different fermentation processes must be monitored, such as varying formulations in craft breweries. In this work, three methodologies are evaluated to use previously collected ultrasonic sensor data from laboratory scale fermentations to improve machine learning model accuracy on an industrial scale fermentation process. These methodologies include training models on both domains simultaneously, training models in a federated learning strategy to preserve data privacy, and fine-tuning the best performing models on the industrial scale data. All methodologies provided increased prediction accuracy compared with training based solely on the industrial fermentation data. The federated learning methodology performed best, achieving higher accuracy for 14 out of 16 machine learning tasks compared with the base case model.

show abstract

Section: Machine Learningcontrasting

confidence: 63%

Domain Adaptation and Federated Learning for Ultrasonic Monitoring of Beer Fermentation

2021

View full text Add to dashboard Cite

show abstract

“…The final model is again distributed to devices to get inference related results. 46 The difference between distributed learning and federated learning is that in federated learning training is initialized on local epochs.…”

Section: Federated Learningmentioning

confidence: 99%

A novel game theory based reliable proof‐of‐stake consensus mechanism for blockchain

Bala

Kaur

2022

Trans Emerging Tel Tech

View full text Add to dashboard Cite

The prominent achievement of blockchain technology stimulates exceptional innovation. The major component of blockchain is the consensus mechanism. The standard consensus mechanisms specifically Proof‐of‐Work (PoW) rely on mining procedures and stake‐based mechanisms such as Proof‐of‐Stake (PoS) rely on massive stake investment as the sole criteria for selection of leader nodes. However, PoW impose huge computational power requirements and latter may incorporate malicious nodes as leader nodes in anonymous blockchain. These issues might fuel the way for distrust among the participants in blockchain. Henceforth, a novel game theory based reliable PoS mechanism for blockchain has been proposed. Federated learning has been used to compute trust_score for each node. The nodes are trained on locally generated dataset. Further, a game theoretic approach has been proposed that uses a reward and punishment scheme to ensure threshold level of trust_score maintenance by each node. Finally, a crop insurance use case has been developed with the consensus mechanism and blockchain coded in python. The insurance claims are made to operate through smart contract based mobile app system to impart more authenticity. The system is tested and results show an intrusion accomplishment rate reduced by approximate 40% when compared to the standard PoS mechanism and by approximately 33% for algorand, 29% for ouroboros and 20% for tendermint. The mean absolute error also decreases by 30% within specific time. Furthermore, the proposed federated learning‐based system is compared with basic neural network‐based machine learning model and the results reveal that a significant reduction in average training time amounting to 8.35 second is achieved. Test accuracy has also been analyzed for various learning mechanisms.

show abstract

“…As for the hidden neuron number in progressive ELM and S-ELM, it is set to 250. Activation function is set to ReLU according to [9].…”

Section: B Experimental Settingsmentioning

confidence: 99%

A Scalable Extreme Learning Machine (S-ELM) for Class-Incremental ECG-Based User Identification

Lee

Chen

2021

2021 IEEE International Symposium on Circuits and Systems (ISCAS)

Self Cite

View full text Add to dashboard Cite

User identification using electrocardiogram (ECG) is emerging due to the uniqueness and convenience of ECG signals.In addition, in real world applications, new subjects may enter the existed identification system and be authorized to access the private data. Therefore, we propose a scalable extreme learning machine (S-ELM) to meet the demand for class-incremental ECGbased user identification. We first prove that the output weight of the S-ELM learnt in a class-incremental manner is as same as that of a regular ELM which prior has the information of the number of total class. Therefore, in our experiment, S-ELM immunes from the catastrophic forgetting phenomenon, which is a common problem in class-incremental scenarios. Comparing to another class-incremental extreme learning machine such as progressive ELM, S-ELM outperforms progressive ELM by 7% accuracy in online dataset. Comparing to another commonly applied classifier, support vector machine (SVM) with linear and radial basis function (RBF) kernels, S-ELM shows its efficiency by 13.35% and 10.54% higher accuracy but only spends 5.09% and 3.48% of the inference time. Therefore, the proposed S-ELM is promising for the class-incremental ECG-based user identification.

show abstract

Online Extreme Learning Machine Design for the Application of Federated Learning

Cited by 12 publications

References 6 publications

Domain Adaptation and Federated Learning for Ultrasonic Monitoring of Beer Fermentation

Domain Adaptation and Federated Learning for Ultrasonic Monitoring of Beer Fermentation

A novel game theory based reliable proof‐of‐stake consensus mechanism for blockchain

A Scalable Extreme Learning Machine (S-ELM) for Class-Incremental ECG-Based User Identification

Contact Info

Product

Resources

About