scholarly journals Multiscale structural complexity of natural patterns

2020 ◽  
Vol 117 (48) ◽  
pp. 30241-30251
Author(s):  
Andrey A. Bagrov ◽  
Ilia A. Iakovlev ◽  
Askar A. Iliasov ◽  
Mikhail I. Katsnelson ◽  
Vladimir V. Mazurenko
Keyword(s):  
Three Dimensional ◽  
Structural Complexity ◽  
Machine Learning Techniques ◽  
Nonequilibrium Systems ◽  
Machine Method ◽  
Standard Methods ◽  
Universal Machine ◽  
Learning Techniques ◽  
Single Number ◽  
Computing Correlation

Complexity of patterns is key information for human brain to differ objects of about the same size and shape. Like other innate human senses, the complexity perception cannot be easily quantified. We propose a transparent and universal machine method for estimating structural (effective) complexity of two-dimensional and three-dimensional patterns that can be straightforwardly generalized onto other classes of objects. It is based on multistep renormalization of the pattern of interest and computing the overlap between neighboring renormalized layers. This way, we can define a single number characterizing the structural complexity of an object. We apply this definition to quantify complexity of various magnetic patterns and demonstrate that not only does it reflect the intuitive feeling of what is “complex” and what is “simple” but also, can be used to accurately detect different phase transitions and gain information about dynamics of nonequilibrium systems. When employed for that, the proposed scheme is much simpler and numerically cheaper than the standard methods based on computing correlation functions or using machine learning techniques.

NIRS Calibration of Aflatoxin in Maize

2018 ◽  
Vol 71 (11) ◽  
pp. 868 ◽  
Author(s):  
Ross E. Darnell ◽  
Jagger J. Harvey ◽  
Glen P. Fox ◽  
Mary T. Fletcher ◽  
James Wainaina ◽  
...  
Keyword(s):  
Near Infrared ◽  
False Negative ◽  
Field Trials ◽  
Aflatoxin Contamination ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Machine Method ◽  
Breeding Programs ◽  
Wet Chemistry ◽  
Learning Techniques

The aim of this study is to determine the value of near-infrared spectroscopy (NIRS) as a diagnostic tool for aflatoxin contamination, specifically to rapidly predict levels of aflatoxin, either quantitatively or qualitatively, in ground maize. Maize was collected from inoculated field trials conducted across four sites in Kenya. Inoculated and uninoculated maize ears were harvested, milled, and prepared for NIRS scanning and wet chemistry-based aflatoxin quantification. Several statistical and machine learning techniques were compared. Absorbance at a single bandwidth explained 34 % of the variation in levels of aflatoxin using a regression model while a partial least-squares (PLS) method showed that NIR measurements could explain 42 % of the variation in aflatoxin levels. To compare various methods for their ability to classify samples with high (>100 ppb) levels of aflatoxin, various additional procedures were applied. The k-nearest neighbour classification method yielded sensitivity and specificity values of 0.75 and 0.52 respectively, compared with the support vector machine method with estimates of 0.81 and 0.68, whereas PLS could achieve values of 0.82 and 0.72 respectively. The corresponding false positive and false negative values are still unacceptable for NIRS to be used with confidence, as ~18 % of contaminated ground maize samples would be accepted and 28 % of good maize would be discarded or declared contaminated or downgraded. However, such calibrations could be useful in breeding programs without access to wet chemistry analysis, seeking to rank entries semiquantitatively.

scholarly journals Geomorphometry today

2021 ◽  
Vol 27 (2) ◽  
pp. 394-448
Author(s):  
Igor Florinsky
Keyword(s):  
Noise Suppression ◽  
Three Dimensional ◽  
Aerial Survey ◽  
Machine Learning Techniques ◽  
Multi Scale ◽  
Learning Techniques ◽  
Digital Elevation ◽  
Analytical Review ◽  
The Subject

Topography is the most important component of the geographical shell, one of the main elements of geosystems, and the framework of a landscape. geomorphometry is a science, the subject of which is modeling and analyzing the topography and the relationships between topography and other components of geosystems. Currently, the apparatus of geomorphometry is widely used to solve various multi-scale problems of the Earth sciences. As part of the RFBR competition “Expansion”, we present an analytical review of the development of theory, methods, and applications of geomorphometry for the period of 2016–2021. For the analysis, we used a sample of 485 of the strongest and most original papers published in international journals belonging to the JCR Web of Science Core Collection quartile I and II (Q1–Q2), as well as monographs from leading international publishers. We analyze factors caused a progress in geomorphometry in recent years. These include widespread use of unmanned aerial survey and digital photogrammetry, development of tools and methods for survey of submarine topography, emergence of new publicly available digital elevation models (DEMs), development of new methods of DEM preprocessing for their filtering and noise suppression, development of methods of two-dimensional and three-dimensional visualization of DEMs, introduction of machine learning techniques, etc. We consider some aspects of the geomorphometric theory developed in 2016–2021. In particular, a new classification of morphometric values is presented. We discuss new computational methods for calculating morphometric models from DEM, as well as the problems facing the developers and users of such methods. We consider application of geomorphometry for solving multiscale problems of geomorphology, hydrology, soil science, geology, glaciology, speleology, plant science and forestry, zoogeography, oceanology, planetology, landslide studies, remote sensing, urban studies, and archaeology.

scholarly journals Statistics of spatial rotations and projection directions considering molecular symmetry in 3D electron cryo-microscopy

2020 ◽  
Author(s):  
Qi Zhang ◽  
Hai Lin ◽  
Mingxu Hu
Keyword(s):  
Three Dimensional ◽  
Machine Learning Techniques ◽  
Molecular Symmetry ◽  
Domain Specific ◽  
Learning Techniques ◽  
Statistical Justification ◽  
Projection Images ◽  
The Mean ◽  
Spatial Rotations ◽  
Unique Games

AbstractElectron cryo-microscopy (cryoEM) three-dimensional (3D) reconstruction is based on estimations of orientations of projection images or 3D volumes. It is common that the macromolecules studied by cryoEM have molecular symmetry, which, unfortunately, has not been taken into consideration by any statistics for either spatial rotations or projection directions at this point. Meanwhile, there are growing needs to adopt advanced statistical methods, and further, modern machine learning techniques in cryoEM. Since those methodologies are built heavily upon statistical learning cornerstones, the absence of their domain-specific statistical justification limits their applications in cryoEM. In this research, based on the concept of non-unique-games (NUG), we propose two key statistical measurements, the mean and the variance, of both spatial rotations and projection directions when molecular symmetry is considered. Such methods are implemented in the open-source python package pySymStat.

scholarly journals Alzheimer's Disease: A Survey

2021 ◽  
Vol 8 (1) ◽  
pp. 33-39
Author(s):  
Harshitha ◽  
Gowthami Chamarajan ◽  
Charishma Y
Keyword(s):  
Alzheimer’S Disease ◽  
Alzheimer's Disease ◽  
Deep Neural Networks ◽  
Brain Area ◽  
Brain Mri ◽  
Three Dimensional ◽  
The Other ◽  
Machine Learning Techniques ◽  
Learning Techniques ◽  
The Brain

Alzheimer's Diseases (AD) is one of the type of dementia. This is one of the harmful disease which can lead to death and yet there is no treatment. There is no current technique which is 100% accurate for the treatment of this disease. In recent years, Neuroimaging combined with machine learning techniques have been used for detection of Alzheimer's disease. Based on our survey we came across many methods like Convolution Neural Network (CNN) where in each brain area is been split into small three dimensional patches which acts as input samples for CNN. The other method used was Deep Neural Networks (DNN) where the brain MRI images are segmented to extract the brain chambers and then features are extracted from the segmented area. There are many such methods which can be used for detection of Alzheimer’s Disease.

Modeling of snow avalanche dynamics using open source software OpenFOAM

2021 ◽  
Author(s):  
Daria Romanova ◽  
Margarita Egiit
Keyword(s):  
Three Dimensional ◽  
Snow Avalanche ◽  
Machine Learning Techniques ◽  
Turbulent Regime ◽  
Snow Avalanches ◽  
Advantages And Disadvantages ◽  
Three Dimensional Modeling ◽  
The Neural Network ◽  
Learning Techniques ◽  
Protective Structures

<p>The work is devoted to the comparison of different approaches for modeling the dynamics of dense and powder snow avalanches. Various 3D and 2D approaches are considered. The accuracy of determining the avalanche run-out zone, the interaction of the flow with obstacles, the front speed, and various distributed parameters are evaluated. As objects for comparison, an experiment on the interaction of a slushflow with a combination of protective structures and a powder snow avalanche in the Khibiny mountains are modeled.</p><p> </p><p>Taking into account the advantages and disadvantages of various approaches based on basic solutions available in the OpenFOAM package, a specialized software avalancheFoam is being developed for three-dimensional modeling of the dynamics of snow avalanches, taking into account the complex turbulent regime and multiphase structure of the flow. Machine learning techniques are used to refine turbulent stresses. The neural network is trained on a dataset obtained from high-precision supercomputer simulation of the flow, and sets the form of additional refinement members of the mathematical model of less computational complexity. Various avalanche sites in the Khibiny mountains are modeled to validate the developed software.</p>

MACHINE APPROACHES TO PROTEIN FEATURE PREDICTION

1992 ◽  
Vol 03 (supp01) ◽  
pp. 183-193 ◽  
Author(s):  
David K. Tcheng ◽  
Shankar Subramaniam
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
Protein Structure ◽  
Homology Modeling ◽  
Three Dimensional ◽  
Antibody Fragments ◽  
Machine Learning Techniques ◽  
Knowledge Based ◽  
Learning Techniques ◽  
Protein Feature

Knowledge-based approaches are being increasingly used in predicting protein structure and motifs. Machine learning techniques such as neural networks and decision-trees have become invaluable tools for these approaches. This paper describes the use of machine learning in predicting sequence-based motifs in antibody fragments. Given the limited number of three dimensional structures and the plethora of sequences, this technique is useful for homology modeling of three dimensional structures of antibody fragments.

scholarly journals Predictive Analysis Of Breast Cancer Using Machine Learning Techniques

2019 ◽  
Vol 15 (29) ◽  
pp. 1-23 ◽  
Author(s):  
Rashmi Agrawal
Keyword(s):  
Breast Cancer ◽  
Machine Learning ◽  
Data Mining ◽  
Machine Learning Techniques ◽  
Predictive Analysis ◽  
Support Vector ◽  
International Institute ◽  
Machine Method ◽  
Linear Discriminant ◽  
Learning Techniques

This paper is a product of the research Project “Predictive Analysis Of Breast Cancer Using Machine Learning Techniques” performed in Manav Rachna International Institute of Research and Studies, Faridabad in the year 2018. Introduction: The present article is part of the effort to predict breast cancer which is a serious concern for women’s health. Problem: Breast cancer is the most common type of cancer and has always been a threat to women’s lives. Early diagnosis requires an effective method to predict cancer to allow physicians to distinguish benign and malicious cancer. Researchers and scientists have been trying hard to find innovative methods to predict cancer. Objective: The objective of this paper will be predictive analysis of breast cancer using various machine learning techniques like Naïve Bayes method, Linear Discriminant Analysis, K-Nearest Neighbors and Support Vector Machine method.  Methodology: Predictive data mining has become an instrument for scientists and researchers in the medical field. Predicting breast cancer at an early stage helps in better cure and treatment. KDD (Knowledge Discovery in Databases) is one of the most popular data mining methods used by medical researchers to identify the patterns and the relationship between variables and also helps in predicting the outcome of the disease based upon historical data of datasets. Results: To select the best model for cancer prediction, accuracy of all models will be estimated and the best model will be selected. Conclusion: This work seeks to predict the best technique with highest accuracy for breast cancer. Originality: This research has been performed using R and the dataset taken from UCI machine learning repository. Limitations: The lack of exact information provided by data.

3-D reconstruction of molecular shape from microcrystal thin sections

1983 ◽  
Vol 41 ◽  
pp. 748-749
Author(s):  
S. Cusack ◽  
J.-C. Jésior
Keyword(s):  
Three Dimensional ◽  
Molecular Shape ◽  
Biological Molecules ◽  
Fourier Space ◽  
Single Particles ◽  
Thin Sections ◽  
Standard Methods ◽  
X Ray ◽  
X Ray Crystallography ◽  
Dimensional Reconstruction

Three-dimensional reconstruction techniques using electron microscopy have been principally developed for application to 2-D arrays (i.e. monolayers) of biological molecules and symmetrical single particles (e.g. helical viruses). However many biological molecules that crystallise form multilayered microcrystals which are unsuitable for study by either the standard methods of 3-D reconstruction or, because of their size, by X-ray crystallography. The grid sectioning technique enables a number of different projections of such microcrystals to be obtained in well defined directions (e.g. parallel to crystal axes) and poses the problem of how best these projections can be used to reconstruct the packing and shape of the molecules forming the microcrystal.Given sufficient projections there may be enough information to do a crystallographic reconstruction in Fourier space. We however have considered the situation where only a limited number of projections are available, as for example in the case of catalase platelets where three orthogonal and two diagonal projections have been obtained (Fig. 1).

Using machine learning techniques to reduce data annotation time

2006 ◽  
Author(s):  
Christopher Schreiner ◽  
Kari Torkkola ◽  
Mike Gardner ◽  
Keshu Zhang
Keyword(s):  
Machine Learning ◽  
Machine Learning Techniques ◽  
Data Annotation ◽  
Learning Techniques

Using Machine Learning Algorithms on Prediction of Stock Price

2020 ◽  
Vol 12 (2) ◽  
pp. 84-99
Author(s):  
Li-Pang Chen
Keyword(s):  
Machine Learning ◽  
Stock Price ◽  
Short Term Memory ◽  
Machine Learning Algorithms ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Short Term ◽  
Learning Techniques ◽  
Historical Database ◽  
Long Short Term Memory

In this paper, we investigate analysis and prediction of the time-dependent data. We focus our attention on four different stocks are selected from Yahoo Finance historical database. To build up models and predict the future stock price, we consider three different machine learning techniques including Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN) and Support Vector Regression (SVR). By treating close price, open price, daily low, daily high, adjusted close price, and volume of trades as predictors in machine learning methods, it can be shown that the prediction accuracy is improved.

Sign in / Sign up

Export Citation Format

Share Document