| 1. |
- Pahikkala, Tapio, et al.
(författare)
-
A Parallel Online Regularized Least-squares Machine Learning Algorithm for Future Multi-core Processors.
- 2011
-
Ingår i: PECCS 2011 - Proceedings of the 1st International Conference on Pervasive and Embedded Computing and Communication Systems 2011. - 9789898425485 ; , s. 590-599
-
Konferensbidrag (refereegranskat)abstract
- In this paper we introduce a machine learning system based on parallel online regularized least-squares learning algorithm implemented on a network on chip (NoC) hardware architecture. The system is specifically suitable for use in real-time adaptive systems due to the following properties it fulfills. Firstly, the system is able to learn in online fashion, a property required in almost all real-life applications of embedded machine learning systems. Secondly, in order to guarantee real-time response in embedded multi-core computer architectures, the learning system is parallelized and able to operate with a limited amount of computational and memory resources. Thirdly, the system can learn to predict several labels simultaneously which is beneficial, for example, in multi-class and multi-label classification as well as in more general forms of multi-task learning. We evaluate the performance of our algorithm from 1 thread to 4 threads, in a quad-core platform. A Network-on-Chip platform is chosen to implement the algorithm in 16 threads. The NoC consists of a 4×4 mesh. Results show that the system is able to learn with minimal computational requirements, and that the parallelization of the learning process considerably reduces the required processing time.
|
|
| 2. |
- Hosseinpour, Farhoud, et al.
(författare)
-
A Resource Management Model for Distributed Multi-Task Applications in Fog Computing Networks
- 2021
-
Ingår i: IEEE Access. - : Institute of Electrical and Electronics Engineers (IEEE). - 2169-3536. ; 9, s. 152792-152802
-
Tidskriftsartikel (refereegranskat)abstract
- While the effectiveness of fog computing in Internet of Things (IoT) applications has been widely investigated in various studies, there is still a lack of techniques to efficiently utilize the computing resources in a fog platform to maximize Quality of Service (QoS) and Quality of Experience (QoE). This paper presents a resource management model for service placement of distributed multitasking applications in fog computing through mathematical modeling of such a platform. Our main design goal is to reduce communication between the candidate nodes hosting different task modules of an application by selecting a group of nodes near each other and as close to the source of the data as possible. We propose a method based on a greedy principle that demonstrates a highly scalable and near-optimal performance for resource mapping problems for multitasking applications in fog computing networks. Compared with the commercial Gurobi optimizer, our proposed algorithm provides a mapping solution that obtains 93% of the performance, attributed to a higher communication cost, while outperforming the reference method in terms of the computing speed, cutting the mapping execution time to less than 1% of that of the Gurobi optimizer.
|
|
| 3. |
- Högmander, Milla, et al.
(författare)
-
Luminometric Label Array for Counting and Differentiation of Bacteria
- 2017
-
Ingår i: Analytical Chemistry. - : American Chemical Society (ACS). - 0003-2700 .- 1520-6882. ; 89:5, s. 3208-3216
-
Tidskriftsartikel (refereegranskat)abstract
- Methods for simple and fast detection and differentiation of bacterial species are required, for instance, in medicine, water quality monitoring, and the food industry. Here, we have developed a novel label array method for the counting and differentiation of bacterial species. This method is based on the nonspecific interactions of multiple unstable lanthanide chelates and selected chemicals within the sample leading to a luminescence signal profile that is unique to the bacterial species. It is simple, cost-effective, and/or user-friendly compared to many existing methods, such as plate counts on selective media, automatic (hemocytometer-based) cell counters, flow cytometry, and polymerase chain reaction (PCR)-based methods for identification. The performance of the method was demonstrated with nine single strains of bacteria in pure culture. The limit of detection for counting was below 1000 bacteria per mL, with an average coefficient of variation of 10% achieved with the developed label array. A predictive model was trained with the measured luminescence signals and its ability to differentiate all tested bacterial species from each other, including members of the same genus Bacillus licheniformis and Bacillus subtilis, was confirmed via leave-one-out cross-validation. The suitability of the method for analysis of mixtures of bacterial species was shown with ternary mixtures of Bacillus licheniformis, Escherichia coli JM109, and Lactobacillus reuteri ATCC PTA 4659. The potential future application of the method could be monitoring for contamination in pure cultures; analysis of mixed bacterial cultures, where examining one species in the presence of another could inform industrial microbial processes; and the analysis of bacterial biofilms, where nonspecific methods based on physical and chemical characteristics are required instead of methods specific to individual bacterial species.
|
|
| 4. |
- Sahebi, Golnaz, et al.
(författare)
-
GeFeS : A generalized wrapper feature selection approach for optimizing classification performance
- 2020
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 125
-
Tidskriftsartikel (refereegranskat)abstract
- In this paper, we propose a generalized wrapper-based feature selection, called GeFeS, which is based on a parallel new intelligent genetic algorithm (GA). The proposed GeFeS works properly under different numerical dataset dimensions and sizes, carefully tries to avoid overfitting and significantly enhances classification accuracy. To make the GA more accurate, robust and intelligent, we have proposed a new operator for features weighting, improved the mutation and crossover operators, and integrated nested cross-validation into the GA process to properly validate the learning model. The k-nearest neighbor (kNN) classifier is utilized to evaluate the goodness of selected features. We have evaluated the efficiency of GeFeS on various datasets selected from the UCI machine learning repository. The performance is compared with state-of-the-art classification and feature selection methods. The results demonstrate that GeFeS can significantly generalize the proposed multi-population intelligent genetic algorithm under different sizes of two-class and multi-class datasets. We have achieved the average classification accuracy of 95.83%, 97.62%, 99.02%, 98.51%, and 94.28% while reducing the number of features from 56 to 28, 34 to 18, 279 to 135, 30 to 16, and 19 to 9 under lung cancer, dermatology, arrhythmia, WDBC, and hepatitis, respectively.
|
|
