TRAINING DATA SAMPLING FOR CONVENTIONAL NEURAL NETWORKS CONFIGURING
نویسندگان
چکیده
منابع مشابه
Training Neural Networks with Deficient Data
We analyze how data with uncertain or missing input features can be incorporated into the training of a neural network. The general solution requires a weighted integration over the unknown or uncertain input although computationally cheaper closed-form solutions can be found for certain Gaussian Basis Function (GBF) networks. We also discuss cases in which heuristical solutions such as substit...
متن کاملTraining Neural Networks with Deecient Data
We analyze how data with uncertain or missing input features can be incorporated into the training of a neural network. The general solution requires a weighted integration over the unknown or uncertain input although computationally cheaper closed-form solutions can be found for certain Gaussian Basis Function (GBF) networks. We also discuss cases in which heuristical solutions such as substit...
متن کاملTraining Neural Networks Using Ever-increasing Data
Peidong Wang Abstract This essay proposed a solution to train artificial neural networks using ever-increasing data. Its background is that in real world applications, training data is accumulated continuously. Current network structures and training methods have to use all of the available data, including data collected previously, to retrain a network. Providing an innovative perspective of t...
متن کاملConfiguring Spiking Neural Networks for Given Spatio-Temporal Patterns
We developed a general framework to configure a spiking neuronal network so that it can precisely generate a desired spatio-temporal pattern of spikes. The unit of spiking neuronal networks employed here is a leaky integrate-and-fire model. Robustness of configured spiking neuronal network is discussed, which leads us to use some routine methods in linear-programming to solve the set of inequal...
متن کاملSobolev Training for Neural Networks
At the heart of deep learning we aim to use neural networks as function approximators – training them to produce outputs from inputs in emulation of a ground truth function or data creation process. In many cases we only have access to input-output pairs from the ground truth, however it is becoming more common to have access to derivatives of the target output with respect to the input – for e...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Electronics and Control Systems
سال: 2020
ISSN: 1990-5548
DOI: 10.18372/1990-5548.66.15225