Filters








10,944 Hits in 9.8 sec

A Bayesian compressed-sensing approach for reconstructing neural connectivity from subsampled anatomical data

Yuriy Mishchenko, Liam Paninski
2012 Journal of Computational Neuroscience  
We describe the possible design of these experiments and develop a Bayesian framework for extracting neural connectivity from compilations of such data.  ...  As a concrete example, we consider a realistic hypothetical connectivity reconstruction experiment in C. elegans, a popular neuroscience model where a complete wiring diagram has been previously obtained  ...  David Hall for his re-examination of the original print data from (White et al., 1986) , and for providing the statistics about the spatial distribution of synapses in C. elegans used in section 3.4.  ... 
doi:10.1007/s10827-012-0390-z pmid:22437567 fatcat:tzqmp6btnnfcdhd7k4oq2ekwui

A subsampling approach for Bayesian model selection [article]

Jon Lachmann and Geir Storvik and Florian Frommlet and Aliaksadr Hubin
2022 arXiv   pre-print
To address this problem, we suggest using a version of a popular batch stochastic gradient descent (BSGD) algorithm for estimating the marginal likelihood of a GLM by subsampling from the data.  ...  This allows performing Bayesian model selection and model averaging.  ...  SGD has received a lot of recent attention in the machine learning community where it is the workhorse of neural networks among other techniques.  ... 
arXiv:2201.13198v1 fatcat:mb7ihwmigratjpjqtzqbqfmpke

Imputing manufacturing material in data mining

Ruey-Ling Yeh, Ching Liu, Ben-Chang Shia, Yu-Ting Cheng, Ya-Fang Huwang
2007 Journal of Intelligent Manufacturing  
When the missing data is continuous, regression models and Neural Networks are used to build imputative models.  ...  For the categorical missing data, the logistic regression model, neural network, C5.0 and CART are employed to construct imputative models.  ...  In the case of the application of regression and neural network model to missing data where the data are of continuous, the result from the simulation clearly indicates better model performance from regression  ... 
doi:10.1007/s10845-007-0067-z fatcat:vhuymsohxbforapwcf2pf77mpe

Data mining for assessing the credit risk of local government units in Croatia

Silvija Vlah Jerić, Marko Primorac
2017 Croatian Operational Research Review  
This paper aims to test the performance of three methods: (1) an artificial neural network (ANN), (2) a hybrid artificial neural network and genetic algorithm approach (ANN-GA), and (2) the Tobit regression  ...  Over the past few decades, data mining techniques, especially artificial neural networks, have been used for modelling many real-world problems.  ...  The Tobit regression approach Analyzing a local unit's credit risk will rely on using a censored regression model, i.e., the Tobit model [36] as suggested in work of [27] .  ... 
doi:10.17535/crorr.2017.0012 fatcat:pegd7mmvtfdy3da6uh7qszk6aq

Feature Selection on Lyme Disease Patient Survey Data [article]

Joshua Vendrow, Jamie Haddock, Deanna Needell, Lorraine Johnson
2020 arXiv   pre-print
We use basic linear regression, support vector machines, neural networks, entropy-based decision tree models, and k-nearest neighbors approaches.  ...  Lyme disease is a rapidly growing illness that remains poorly understood within the medical community.  ...  The authors would like to thank LymeDisease.org for the use of data derived from the MyLymeData patient registry, Phase 1 27 April 2017.  ... 
arXiv:2009.09087v1 fatcat:kvihqymcabeetluv27bfvkeyve

Estimating population proportions from imputed data

Svein Nordbotten
1998 Computational Statistics & Data Analysis  
Imputation estimates based on imputed values obtained from neural network models used in an 'impute first-aggregate next' approach, have been computed from Norwegian population census and administrative  ...  The imputation estimates were compared with simple unbiased estimates obtained by the traditional 'aggregate firstestimate next' approach and found to be superior for estimating proportions in small subgroups  ...  For training samples as large as the one used in the experiments, the sampling should not be expected to have a significant influence on the results.  ... 
doi:10.1016/s0167-9473(98)00011-5 fatcat:ikq5m7t67zgexhbmskci6sexvm

Forest Attributes Estimation Using Aerial Laser Scanner and TM Data

S. Shataee Joibary
2013 Forest Systems  
Area of study: Data in small part of a mixed managed forest in the Waldkirch region, Germany.  ...  and ANN with different types of networks.  ...  in my sabatical duration.  ... 
doi:10.5424/fs/2013223-03874 fatcat:nof4lejlhbfozomz2facen2xma

Machine Learning for Financial Stability [chapter]

Lucia Alessi, Roberto Savona
2021 Data Science for Economics and Finance  
A strand of literature has shown that machine learning approaches can make more accurate data-driven predictions than standard empirical models, thus providing more and more timely information about the  ...  However, as "black box" models, they are still much underutilized in financial stability, a field where interpretability and accountability are crucial.  ...  Finally, Unsupervised Artificial Neural Networks (U-ANN) are used when dealing with unlabeled data sets.  ... 
doi:10.1007/978-3-030-66891-4_4 fatcat:cxngfuhchzbqniql72uvtlcvfm

Training the Convolutional Neural Network with Statistical Dependence of the Response on the Input Data Distortion [article]

Igor Janiszewski, Dmitry Slugin, Vladimir V. Arlazarov
2019 arXiv   pre-print
As an example, the LeNet5 architecture network with training data based on the MNIST symbols and a distortion model as Gaussian blur with a variable level of distortion is considered.  ...  The paper proposes an approach to training a convolutional neural network using information on the level of distortion of input data.  ...  Because of that only samples with ≥ 0.5 was used for training. EXPERIMENT RESULTS Consider the convolutional neural network like LeNet5 [10] with training on the selected data model.  ... 
arXiv:1912.00664v1 fatcat:auyhowshtnaznpvbts5dkgzfue

LogitBoost autoregressive networks

Marc Goessling
2017 Computational Statistics & Data Analysis  
Recently popular approaches have modeled these conditionals through neural networks with sophisticated weight-sharing structures.  ...  Similarities and differences between the proposed approach and autoregressive models based on neural networks are discussed in detail.  ...  We propose LogitBoost as a learning procedure for the conditionals in an autoregressive network. In existing work (Shafik and Tutz, 2009) boosting was applied to continuous, stationary time series.  ... 
doi:10.1016/j.csda.2017.03.010 fatcat:xmmvrf3nqjbv3etiymz3d33hae

Data-driven forecasting of solar irradiance [article]

Pierrick Bruneau and Philippe Pinheiro and Yoann Didry
2019 arXiv   pre-print
techniques such as neural networks and regression trees.  ...  After describing our data cleaning and normalization process, we combine a variable selection step based on AutoRegressive Integrated Moving Average (ARIMA) models, to using general purpose regression  ...  A Bayesian selection method has been proposed for MLP models in [5] , but this requires initial training of neural networks with deliberately large input vectors, over which selection is performed a posteriori  ... 
arXiv:1801.03373v2 fatcat:xbjwiigquzdlbio7hj3ecj7y4q

A general, simple, robust method to account for measurement error when analyzing data with an internal validation subsample [article]

Walter K Kremers
2021 arXiv   pre-print
Methods: Here we describe a general method accounting for measurement error in outcome and/or predictor variables for the parametric regression setting when there is a validation subsample where variables  ...  This method should be a valuable tool in the analysis of data with measurement error.  ...  error in the validation subsample, and one example with patient data.  ... 
arXiv:2106.14063v2 fatcat:azzlfiebqzhidjnhdwxojzoxdu

History and Potential of Binary Segmentation for Exploratory Data Analysis

James N. Morgan
2021 Journal of Data Science  
Where there is a clear criterion (dependent) variable or classification, sequential binary segmentation (tree) programs are being used.  ...  Exploratory data analysis has become more important as large rich data sets become available, with many explanatory variables representing competing theoretical constructs.  ...  A variety of data-mining or neural network programs is also on the market, with apparently different objectives, though the explanations of what they do leave much to be desired (see http://www.kdnuggets.com  ... 
doi:10.6339/jds.2005.03(2).198 fatcat:z35y3m6llnhypcuwydttosfe2e

A Vision Inspired Neural Network for Unsupervised Anomaly Detection in Unordered Data [article]

Nassir Mohammad
2022 arXiv   pre-print
A robust and accurate model for anomaly detection in univariate and multivariate data is built using this network as a concrete application.  ...  More formally anomalies are those observations-under appropriate random variable modelling-whose expectation of occurrence with respect to a grouping of prior interest is less than one; such a definition  ...  Similarly the neuron model has thresholds but which are data dependent (adaptive), inherently dynamic and selective to the output.  ... 
arXiv:2205.06716v1 fatcat:nlukbpifffbvdfxv365jwoe254

Data Driven Performance Prediction in Steel Making

Fernando Boto, Maialen Murua, Teresa Gutierrez, Sara Casado, Ana Carrillo, Asier Arteaga
2022 Metals  
A new approach based on ensembles has been developed with feature selection methods and four state-of-the-art regression approximations (random forest, gradient boosting, xgboost and neural networks).  ...  This work presents three data-driven models based on process data, to estimate different indicators related to process performance in a steel production process.  ...  The selected regression modeling approaches are largely based on ensemble learning and a neural network has been included, as it is the most widely used strategy in the literature.  ... 
doi:10.3390/met12020172 fatcat:wmmsk3b75zhxxnxjni5x6cjgi4
« Previous Showing results 1 — 15 out of 10,944 results