http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Hybrid no-propagation learning for multilayer neural networks
Adhikari, Shyam Prasad,Yang, Changju,Slot, Krzysztof,Strzelecki, Michal,Kim, Hyongsuk Elsevier 2018 Neurocomputing Vol.321 No.-
<P><B>Abstract</B></P> <P>A hybrid learning algorithm suitable for hardware implementation of multi-layer neural networks is proposed. Though backpropagation is a powerful learning method for multilayer neural networks, its hardware implementation is difficult due to complexities of the neural synapses and the operations involved in error backpropagation. We propose a learning algorithm with performance comparable to but easier than backpropagation to be implemented in hardware for on-chip learning of multi-layer neural networks. In the proposed learning algorithm, a multilayer neural network is trained with a hybrid of gradient-based delta rule and a stochastic algorithm, called Random Weight Change. The parameters of the output layer are learned using the delta rule, whereas the inner layer parameters are learned using Random Weight Change, thereby the overall multilayer neural network is trained without the need for error backpropagation. Experimental results showing better performance of the proposed hybrid learning rule than either of its constituent learning algorithms, and comparable to that of backpropagation on the benchmark MNIST dataset are presented. Hardware architecture illustrating the ease of implementation of the proposed learning rule in analog hardware vis-a-vis the backpropagation algorithm is also presented.</P>
( Jesse W.P. Kuiper ),( Steven J. Verberne ),( Pim W. van Egmond ),( Karin Slot ),( Olivier P.P. Temmerman ),( Constantijn J. Vos ) 대한고관절학회 2022 Hip and Pelvis Vol.34 No.4
Purpose: The most recent diagnostic criteria for periprosthetic joint infection (PJI) include the use of the alphadefensin (AD) lateral-flow (LF) test, but hip and knee arthroplasties were usually combined in previous studies. This prospective study was designed to examine the accuracy of the AD-LF test for diagnosis of PJI in chronic painful total hip arthroplasties (THA). Materials and Methods: Patients with chronic painful hip arthroplasties were prospectively enrolled between March 2018 and May 2020. Exclusion criteria included acute PJI or an insufficient amount of synovial fluid. The modified Musculoskeletal Infection Society (MSIS) criteria were primarily used for PJI diagnosis. Fifty-seven patients were included in the analysis group. Revision surgery was not performed in 38 patients, for different reasons (clinical group); these patients remain “Schrödinger’s hips”: in such cases PJI cannot be excluded nor confirmed until you “open the box”. Results: The result of the AD-LF test was positive in nine patients and negative in 48 patients. Six patients were diagnosed with PJI. AD-LF sensitivity (MSIS criteria) was 83% (95% confidence interval [CI] 36-100%) and specificity was 92% (95% CI 81-98%). The positive and negative predictive value were 56% and 98%, respectively. Conclusion: The AD test is useful in addition to the existing arsenal of diagnostic tools, and can be helpful in the decision-making process. Not all patients with chronical painful THA will undergo revision surgery. Consequently, in order to determine the reliable diagnostic accuracy of this test, future PJI diagnostic studies should include a second arm of “Schrödinger’s hips”.