scholarly journals Radial and Directional Posteriors for Bayesian Deep Learning

2020 ◽  
Vol 34 (04) ◽  
pp. 5298-5305
Author(s):  
Changyong Oh ◽  
Kamil Adamczewski ◽  
Mijung Park
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Mean Field ◽  
Predictive Performance ◽  
Radial Component ◽  
Field Type ◽  
Input And Output ◽  
Output Neurons ◽  
Directional Density ◽  
Directional Component

We propose a new variational family for Bayesian neural networks. We decompose the variational posterior into two components, where the radial component captures the strength of each neuron in terms of its magnitude; while the directional component captures the statistical dependencies among the weight parameters. The dependencies learned via the directional density provide better modeling performance compared to the widely-used Gaussian mean-field-type variational family. In addition, the strength of input and output neurons learned via our posterior provides a structured way to compress neural networks. Indeed, experiments show that our variational family improves predictive performance and yields compressed networks simultaneously.

scholarly journals Comparing Deep and Shallow Neural Networks in Forecasting Call Center Arrivals

2021 ◽  
Author(s):  
Andrea Manno ◽  
Fabrizio Rossi ◽  
Stefano Smriglio ◽  
Luigi Cerone
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Call Center ◽  
Planning Process ◽  
Workforce Planning ◽  
Predictive Performance ◽  
Learning Method ◽  
Input Selection ◽  
Echo State Networks ◽  
Real World Datasets

Abstract Forecasting volumes of incoming calls is the first step of the workforce planning process in call centers and represents a prominent issue from both research and industry perspectives. We investigate the application of Neural Networks to predict incoming calls 24 hours ahead. In particular, a Deep Learning method known as Echo State Networks, is compared with a completely different shallow Neural Networks strategy, in which the lack of recurrent connections is compensated by a careful input selection. The comparison, carried out on three different real world datasets, reveals similar predictive performance, although the shallow approach seems to be more robust and less demanding in terms of time-to-predict.

Estimating Pigment Concentrations from Spectral Images Using an Encoder‐Decoder Neural Network

2020 ◽  
Vol 64 (3) ◽  
pp. 30502-1-30502-15
Author(s):  
Kensuke Fukumoto ◽  
Norimichi Tsumura ◽  
Roy Berns
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Absorption Coefficient ◽  
Spectral Data ◽  
High Accuracy ◽  
Pigment Concentration ◽  
Scattering Coefficient ◽  
A Value ◽  
Input And Output ◽  
Pigment Concentrations

Abstract A method is proposed to estimate the concentration of pigments mixed in a painting, using the encoder‐decoder model of neural networks. The model is trained to output a value that is the same as its input, and its middle output extracts a certain feature as compressed information about the input. In this instance, the input and output are spectral data of a painting. The model is trained with pigment concentration as the middle output. A dataset containing the scattering coefficient and absorption coefficient of each of 19 pigments was used. The Kubelka‐Munk theory was applied to the coefficients to obtain many patterns of synthetic spectral data, which were used for training. The proposed method was tested using spectral images of 33 paintings, which showed that the method estimates, with high accuracy, the concentrations that have a similar spectrum of the target pigments.

Levenshtein Augmentation Improves Performance of SMILES Based Deep-Learning Synthesis Prediction

2020 ◽  
Author(s):  
Dean Sumner ◽  
Jiazhen He ◽  
Amol Thakkar ◽  
Ola Engkvist ◽  
Esben Jannik Bjerrum
Keyword(s):  
Neural Networks ◽  
Pattern Recognition ◽  
Deep Learning ◽  
Recurrent Neural Networks ◽  
Data Augmentation ◽  
State Of The Art ◽  
Sequence Similarity ◽  
Learning Models ◽  
Underlying Network

<p>SMILES randomization, a form of data augmentation, has previously been shown to increase the performance of deep learning models compared to non-augmented baselines. Here, we propose a novel data augmentation method we call “Levenshtein augmentation” which considers local SMILES sub-sequence similarity between reactants and their respective products when creating training pairs. The performance of Levenshtein augmentation was tested using two state of the art models - transformer and sequence-to-sequence based recurrent neural networks with attention. Levenshtein augmentation demonstrated an increase performance over non-augmented, and conventionally SMILES randomization augmented data when used for training of baseline models. Furthermore, Levenshtein augmentation seemingly results in what we define as <i>attentional gain </i>– an enhancement in the pattern recognition capabilities of the underlying network to molecular motifs.</p>

Predictive performance modeling of Habesha brewery wastewater treatment plant using artificial neural networks

2019 ◽  
Author(s):  
Chem Int
Keyword(s):  
Neural Networks ◽  
Water Quality ◽  
Wastewater Treatment ◽  
Artificial Neural Networks ◽  
Performance Modeling ◽  
Predictive Performance ◽  
Mathematical Tool ◽  
Effluent Water ◽  
Artificial Neural

Recently, process control in wastewater treatment plants (WWTPs) is, mostly accomplished through examining the quality of the water effluent and adjusting the processes through the operator’s experience. This practice is inefficient, costly and slow in control response. A better control of WTPs can be achieved by developing a robust mathematical tool for performance prediction. Due to their high accuracy and quite promising application in the field of engineering, Artificial Neural Networks (ANNs) are attracting attention in the domain of WWTP predictive performance modeling. This work focuses on applying ANN with a feed-forward, back propagation learning paradigm to predict the effluent water quality of the Habesha brewery WTP. Data of influent and effluent water quality covering approximately an 11-month period (May 2016 to March 2017) were used to develop, calibrate and validate the models. The study proves that ANN can predict the effluent water quality parameters with a correlation coefficient (R) between the observed and predicted output values reaching up to 0.969. Model architecture of 3-21-3 for pH and TN, and 1-76-1 for COD were selected as optimum topologies for predicting the Habesha Brewery WTP performance. The linear correlation between predicted and target outputs for the optimal model architectures described above were 0.9201 and 0.9692, respectively.

scholarly journals Deep convolutional neural networks for cardiovascular vulnerable plaque detection

2019 ◽  
Vol 277 ◽  
pp. 02024 ◽  
Author(s):  
Lincan Li ◽  
Tong Jia ◽  
Tianqi Meng ◽  
Yizhe Liu
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Vulnerable Plaque ◽  
Recall Rate ◽  
Superior Performance ◽  
Learning Approaches ◽  
Deep Convolutional Neural Networks ◽  
Vulnerable Plaques ◽  
Plaque Detection

In this paper, an accurate two-stage deep learning method is proposed to detect vulnerable plaques in ultrasonic images of cardiovascular. Firstly, a Fully Convonutional Neural Network (FCN) named U-Net is used to segment the original Intravascular Optical Coherence Tomography (IVOCT) cardiovascular images. We experiment on different threshold values to find the best threshold for removing noise and background in the original images. Secondly, a modified Faster RCNN is adopted to do precise detection. The modified Faster R-CNN utilize six-scale anchors (122,162,322,642,1282,2562) instead of the conventional one scale or three scale approaches. First, we present three problems in cardiovascular vulnerable plaque diagnosis, then we demonstrate how our method solve these problems. The proposed method in this paper apply deep convolutional neural networks to the whole diagnostic procedure. Test results show the Recall rate, Precision rate, IoU (Intersection-over-Union) rate and Total score are 0.94, 0.885, 0.913 and 0.913 respectively, higher than the 1st team of CCCV2017 Cardiovascular OCT Vulnerable Plaque Detection Challenge. AP of the designed Faster RCNN is 83.4%, higher than conventional approaches which use one-scale or three-scale anchors. These results demonstrate the superior performance of our proposed method and the power of deep learning approaches in diagnose cardiovascular vulnerable plaques.

Sign in / Sign up

Export Citation Format

Share Document