Filters








26,018 Hits in 5.6 sec

Discovering Nonlinear Relations with Minimum Predictive Information Regularization [article]

Tailin Wu, Thomas Breuel, Michael Skuhersky, Jan Kautz
2020 arXiv   pre-print
In this work, we introduce a novel minimum predictive information regularization method to infer directional relations from time series, allowing deep learning models to discover nonlinear relations.  ...  Identifying the underlying directional relations from observational time series with nonlinear interactions and complex relational structures is key to a wide range of applications, yet remains a hard  ...  Minimum Predictive Information Regularization (MPIR) method for exploratory discovery of nonlinear directional relations from observational time series.  ... 
arXiv:2001.01885v1 fatcat:wxhcino2zjdu7ewlblv2kzv7sq

Data-driven discovery of coordinates and governing equations

Kathleen Champion, Bethany Lusch, J. Nathan Kutz, Steven L. Brunton
2019 Proceedings of the National Academy of Sciences of the United States of America  
We demonstrate this approach on several example high-dimensional systems with low-dimensional behavior.  ...  In this work, we design a custom deep autoencoder network to discover a coordinate transformation into a reduced space where the dynamics may be sparsely represented.  ...  In related work, Koopman analysis seeks coordinates that linearize nonlinear dynamics (29) ; while linear models are useful for prediction and control, they cannot capture the full behavior of many nonlinear  ... 
doi:10.1073/pnas.1906995116 pmid:31636218 pmcid:PMC6842598 fatcat:m4swjd6e25b77p6bdpwb2mlgxa

Blood Tumor Prediction Using Data Mining Techniques

Alaa M. El-Halees, Asem H. Shurrab
2017 Health Informatics - An International Journal  
In this paper, we applied data mining techniques to discover the relations between blood test characteristics and blood tumor in order to predict the disease in an early stage, which can be used to enhance  ...  Also, it demonstrated that deep learning classifiers has the best ability to predict tumor types of blood diseases with an accuracy of 79.45%.  ...  In our paper, we investigated which CBC test has a relation with blood tumor sample. The second method we used was rule induction. Rule induction discovers patterns hidden in data.  ... 
doi:10.5121/hiij.2017.6202 fatcat:2ejlkq2ganbrfanrk22jhbvdju

Score Prediction of Sports Events Based on Parallel Self-Organizing Nonlinear Neural Network

Junyao Ling, Akshi Kumar
2022 Computational Intelligence and Neuroscience  
When predicting the parallel self-organizing network, the minimum error of the self-organizing difference model is 0.3691, and the minimum error of the self-organizing autoregressive neural network is  ...  First, we train and describe the law and development trend of the parallel self-organizing network through historical data of the parallel self-organizing network and then use the discovered law to predict  ...  deal with uncertain and nonlinear prediction problems, so more research and applications have been obtained in the prediction of chaotic parallel self-organizing networks.  ... 
doi:10.1155/2022/4882309 pmid:35075357 pmcid:PMC8783733 fatcat:qpedtsydbzhhhgjjy5b7yzku7y

Applying nonlinear dynamics features for speech-based fatigue detection

Jarek Krajewski, David Sommer, Thomas Schnupp, Tom Laufenberg, Christian Heinze, Martin Golz
2010 Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - MB '10  
Applying methods of Non Linear Dynamics (NLD) provides additional information regarding the dynamics and structure of fatigue speech comparing to the commonly applied speech emotion recognition feature  ...  Additionally, it would seem beneficial that future studies address the main topics of enriching the steering feature set with easy accessible driving related informations as e.g. pedal movement behaviour  ...  Thus, the aim of this study is to apply nonlinear dynamics (NLD) based features within the field of speech acoustics in order to improve the prediction of fatigue.  ... 
doi:10.1145/1931344.1931367 dblp:conf/mb/KrajewskiSSLHG10 fatcat:sdxdrn7uufadlnpfv2zws6qxpq

Recent Advances in Large Margin Learning [article]

Yiwen Guo, Changshui Zhang
2021 arXiv   pre-print
We managed to shorten the paper such that the crucial spirit of large margin learning and related methods are better emphasized.  ...  This paper serves as a survey of recent advances in large margin training and its theoretical foundations, mostly for (nonlinear) deep neural networks (DNNs) that are probably the most prominent machine  ...  However, it is also discovered that using an identity function for r(·) as − i∈T m xi,f /|T | results in poor prediction accuracy on benign inputs.  ... 
arXiv:2103.13598v2 fatcat:3cqqsiiqa5eptcslipzzyofwpq

Data-driven discovery of coordinates and governing equations [article]

Kathleen Champion, Bethany Lusch, J. Nathan Kutz, Steven L. Brunton
2019 arXiv   pre-print
In this work, we design a custom autoencoder to discover a coordinate transformation into a reduced space where the dynamics may be sparsely represented.  ...  We demonstrate this approach on several example high-dimensional dynamical systems with low-dimensional behavior.  ...  In a related vein, Koopman analysis seeks to discover coordinates that linearize nonlinear dynamics [44] [45] [46] [47] ; while linear models are useful for prediction and control, they cannot capture  ... 
arXiv:1904.02107v2 fatcat:254dtyd2uzbojendrwaxofbibe

Extending Granger causality to nonlinear systems [article]

Nicola Ancona, Daniele Marinazzo, Sebastiano Stramaglia
2004 arXiv   pre-print
Not all the nonlinear prediction schemes are suitable to evaluate causality, indeed not all of them allow to quantify how much the knowledge of the other time series counts to improve prediction error.  ...  We consider extension of Granger causality to nonlinear bivariate time series.  ...  Several papers dealt with this problem relating it to the identification of interdependence in nonlinear dynamical systems [6] , or to estimates of information rates [7, 8] .  ... 
arXiv:physics/0405009v1 fatcat:23qulijmmvhuvnovjlbo4hwdx4

A Hybrid Forecasting Model for Nonstationary and Nonlinear Time Series in the Stochastic Process of CO2 Emission Trading Price Fluctuation

Shanglei Chai, Mo Du, Xi Chen, Wenjun Chu
2020 Mathematical Problems in Engineering  
This model can granulate raw data into fuzzy-information granular components with minimum (Low), average (R), and maximum (Up) values as changing space-description parameters.  ...  nonlinear characteristics.  ...  Predict the Low (Minimum), R (Average), and up (Maximum).  ... 
doi:10.1155/2020/8978504 fatcat:3aaed2rh4jhtjcq4seia4zyqne

Existence of Electrically Charged Structures with Regular Center in Nonlinear Electrodynamics Minimally Coupled to Gravity

Irina Dymnikova, Evgeny Galaktionov, Eduard Tropp
2015 Advances in Mathematical Physics  
We address the question of correct description of Lagrange dynamics for regular electrically charged structures in nonlinear electrodynamics coupled to gravity.  ...  the regular center.  ...  Regular spherically symmetric solutions with the nonzero electric charge [36, [44] [45] [46] [47] were found with using the alternative -form of nonlinear electrodynamics obtained from the standard Lagrangian  ... 
doi:10.1155/2015/496475 fatcat:rel5th7sjzcjrofrnojg2rkrw4

A Nonlinear Autoregressive Exogenous (NARX) Neural Network Model for the Prediction of Timestamp Influence on Bitcoin Value

Nahla Aljojo, Areej Alshutayri, Eman Aldhahri, Seita Almandeel, Azida Zainol
2021 IEEE Access  
That is why this current study utilized a Nonlinear Autoregressive Exogenous (NARX) Neural Network Model for the prediction timestamp influence on Bitcoin value.  ...  Simulation analysis indicates that bitcoin digital currency's performance variation is highly influenced by its transaction timestamp with the prediction accuracy of 96%.  ...  price prediction using regularized deep learning Autoregressive integrative moving average is effective at a minimum reported residual sum of squares of 0.002  ... 
doi:10.1109/access.2021.3124629 fatcat:3hz3yy7r65avng53s5ix4pnup4

Intelligence, physics and information – the tradeoff between accuracy and simplicity in machine learning [article]

Tailin Wu
2020 arXiv   pre-print
Thirdly, how can agents discover causality from observations?  ...  We address part of this question by introducing an algorithm that combines prediction and minimizing information from the input, for exploratory causal discovery from observational time series.  ...  In this chapter, we introduce a novel minimum predictive information regularization method to infer directional relations from time series, allowing deep learning models to discover nonlinear relations  ... 
arXiv:2001.03780v2 fatcat:piduzlhoafcjhhsgthulbbhtke

Gradient Boosted Feature Selection [article]

Zhixiang Eddie Xu, Gao Huang, Kilian Q. Weinberger, Alice X. Zheng
2019 arXiv   pre-print
Yet it scales to larger data set sizes and naturally allows for domain-specific side information.  ...  A feature selection algorithm should ideally satisfy four conditions: reliably extract relevant features; be able to identify non-linear feature interactions; scale linearly with the number of features  ...  [18] propose the Minimum Redundancy Maximum Relevance (mRMR) algorithm, which selects a subset of the most responsive features that have high mutual information with labels.  ... 
arXiv:1901.04055v1 fatcat:eq6mw6f4dndrjeh4s4x5k6vwmu

Gradient boosted feature selection

Zhixiang Xu, Gao Huang, Kilian Q. Weinberger, Alice X. Zheng
2014 Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '14  
Yet it scales to larger data set sizes and naturally allows for domain-specific side information.  ...  A feature selection algorithm should ideally satisfy four conditions: reliably extract relevant features; be able to identify non-linear feature interactions; scale linearly with the number of features  ...  [17] propose the Minimum Redundancy Maximum Relevance (mRMR) algorithm, which selects a subset of the most responsive features that have high mutual information with labels.  ... 
doi:10.1145/2623330.2623635 dblp:conf/kdd/XuHWZ14 fatcat:digaumqncbdf3m2gmadjznmf3q

What needles do sparse neural networks find in nonlinear haystacks [article]

Sylvain Sardy, Nicolas W Hengartner, Nikolai Bonenko, Yen Ting Lin
2020 arXiv   pre-print
Our approach is a generalization of the universal threshold of Donoho and Johnstone (1994) to nonlinear ANN learning.  ...  For linear models, such an approach provably also recovers the important features with high probability in regimes for a well-chosen penalty parameter.  ...  Our contribution is to investigate whether this property extends to nonlinear associations with ANNs to discover their underlying lower-dimensional structures.  ... 
arXiv:2006.04041v1 fatcat:qjywf3uczjhvfa6rf5xovg7lbm
« Previous Showing results 1 — 15 out of 26,018 results