scholarly journals Comparison of Non Linear Models and Artificial Neural Network to Describe the Liveweight from Birth to Maturity in Female Jersey Cattle

2021 ◽  
pp. 114-119
Author(s):  
D. O. Omoniwa ◽  
J. E. T. Akinsola ◽  
R. O. Okeke ◽  
J. M. Madu ◽  
D. S. Bunjah Umar
Keyword(s):  
Neural Network ◽  
Goodness Of Fit ◽  
Inflection Point ◽  
Linear Models ◽  
Network Models ◽  
Linear Functions ◽  
Growth Data ◽  
Neural Network Models ◽  
Jersey Cattle ◽  
Non Linear

Evaluation of growth data is an important strategy to manage gross feed requirement in female Jersey cattle in the New Derived Guinea Savannah Zone of Nigeria. Two non-linear functions (Gompertz and Logistic) and Neural network models were used to fit liveweight (LW)-age data using the non linear procedure of JMP statistical software. Data used for this study were collected from 150 Jersey female cattle in Shonga Dairy Farm, Kwara, State from 2010-2018. The Neural network function showedthe best goodness of fit. Both the Gompertz and Logistic functions overestimated LW at birth, 3, 36, 48, 60 and 72months respectively. NN function overestimated the LW at 0, 3, 24, 36 and 72 months. The Gompertzfunction had the best estimation of asymptotic weight (649.51 kg) with average absolute growth rate (0.061 kg/day).The inflection point was 15.95, 9.55 and 34.5 months in Logistic, Gompertz and neural network models, respectively. A strong and positive correlation was observed between asymptote and inflection point in Gompertz functions. The metrics of goodness of fit criteria (R2 and RMSE), showed that NN with multilayer perceptron was superior to the other models but Gompertz model, was best in its ability to approximate complex functions of growth curve parametersin female Jersey cattle.

scholarly journals Neural network emulation of a rainfall-runoff model

2007 ◽  
Vol 4 (1) ◽  
pp. 287-326 ◽  
Author(s):  
R. J. Abrahart ◽  
L. M. See
Keyword(s):  
Neural Network ◽  
Transfer Functions ◽  
Network Models ◽  
Exact Equation ◽  
Rainfall Runoff ◽  
Neural Network Models ◽  
Linear Relationships ◽  
Non Linear ◽  
Rainfall Runoff Model ◽  
Runoff Model

Abstract. The potential of an artificial neural network to perform simple non-linear hydrological transformations is examined. Four neural network models were developed to emulate different facets of a recognised non-linear hydrological transformation equation that possessed a small number of variables and contained no temporal component. The modeling process was based on a set of uniform random distributions. The cloning operation facilitated a direct comparison with the exact equation-based relationship. It also provided broader information about the power of a neural network to emulate existing equations and model non-linear relationships. Several comparisons with least squares multiple linear regression were performed. The first experiment involved a direct emulation of the Xinanjiang Rainfall-Runoff Model. The next two experiments were designed to assess the competencies of two neural solutions that were developed on a reduced number of inputs. This involved the omission and conflation of previous inputs. The final experiment used derived variables to model intrinsic but otherwise concealed internal relationships that are of hydrological interest. Two recent studies have suggested that neural solutions offer no worthwhile improvements in comparison to traditional weighted linear transfer functions for capturing the non-linear nature of hydrological relationships. Yet such fundamental properties are intrinsic aspects of catchment processes that cannot be excluded or ignored. The results from the four experiments that are reported in this paper are used to challenge the interpretations from these two earlier studies and thus further the debate with regards to the appropriateness of neural networks for hydrological modelling.

scholarly journals Extreme learning machines for weather-based modelling of silk cocoon production

2022 ◽  
Vol 24 (1) ◽  
Author(s):  
PRAMIT PANDIT ◽  
BISHVAJIT BAKSHI ◽  
SHILPA M.
Keyword(s):  
Neural Network ◽  
Network Models ◽  
Extreme Learning Machines ◽  
Linear Regression Models ◽  
Weather Variables ◽  
Neural Network Models ◽  
Cocoon Production ◽  
Silk Cocoon ◽  
Learning Machines ◽  
Non Linear

In spite of the immense popularity and sheer power of the neural network models, their application in sericulture is still very much limited. With this backdrop, this study evaluates the suitability of neural network models in comparison with the linear regression models in predicting silk cocoon production of the selected six districts (Kolar, Chikballapur, Ramanagara, Chamarajanagar, Mandya and Mysuru) of Karnataka by utilising weather variables for ten consecutive years (2009-2018). As the weather variables are found to be correlated, principal components are obtained and fed into the linear (principal component regression) and non-linear models (back propagation-artificial neural network and extreme learning machine) as inputs. Outcomes emanated from this experiment have revealed the clear advantages of employing extreme learning machines (ELMs) for weather-based modelling of silk cocoon production. Application of ELM would be particularly useful, when the relation between production and its attributing characters is complex and non-linear.

scholarly journals Gyro Error Compensation in Optoelectronic Platform Based on a Hybrid ARIMA-Elman Model

Algorithms ◽  
2019 ◽  
Vol 12 (1) ◽  
pp. 22 ◽  
Author(s):  
Xingkui Xu ◽  
Chunfeng Wu ◽  
Qingyu Hou ◽  
Zhigang Fan
Keyword(s):  
Neural Network ◽  
Error Compensation ◽  
Linear Models ◽  
Arima Model ◽  
Network Models ◽  
Random Component ◽  
Neural Network Models ◽  
Angle Sensor ◽  
Elman Neural Network ◽  
Differential Algorithm

As an important angle sensor of the opto-electric platform, gyro output accuracy plays a vital role in the stabilization and track accuracy of the whole system. It is known that the generally used fixed-bandwidth filters, single neural network models, or linear models cannot compensate for gyro error well, and so they cannot meet engineering needs satisfactorily. In this paper, a novel hybrid ARIMA-Elman model is proposed. For the reason that it can fully combine the strong linear approximation capability of the ARIMA model and the superior nonlinear compensation capability of a neural network, the proposed model is suitable for handling gyro error, especially for its non-stationary random component. Then, to solve the problem that the parameters of ARIMA model and the initial weights of the Elman neural network are difficult to determine, a differential algorithm is initially utilized for parameter selection. Compared with other commonly used optimization algorithms (e.g., the traditional least-squares identification method and the genetic algorithm method), the intelligence differential algorithm can overcome the shortcomings of premature convergence and has higher optimization speed and accuracy. In addition, the drift error is obtained based on the technique of lift-wavelet separation and reconstruction, and, in order to weaken the randomness of the data sequence, an ashing operation and Jarque-Bear test have been added to the handle process. In this study, actual gyro data is collected and the experimental results show that the proposed method has higher compensation accuracy and faster network convergence, when compared with other commonly used error-compensation methods. Finally, the hybrid method is used to compensate for gyro error collected in other states. The test results illustrate that the proposed algorithm can effectively improve error compensation accuracy, and has good generalization performance.

scholarly journals A comparison of artificial neural network models and regression models to predict tree volumes for crimean black pine trees in Cankiri forests

2019 ◽  
Vol 143 (9-10) ◽  
pp. 423-423
Author(s):  
Muammer Şenyurt ◽  
Ilker Ercanli
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Basis Function ◽  
Regression Models ◽  
Goodness Of Fit ◽  
Network Models ◽  
Black Pine ◽  
Neural Network Models ◽  
Pine Trees ◽  
Artificial Neural Network Models

Cilj ovog rada je usporediti modele umjetne neuralne mreže (ANN) za predviđanje pojedinih drvnih volumena krimskih borova u šumama Çankirija. Jednoulazne i dvoulazne jednadžbe i kompatibilna volumna jednadžba Fang et al. (2000) temeljena na klasičnim i tradicionalnim metodama primijenjena je na 360 krimskih borova u cilju dobivanja ovih drvnih volumena. Kako bi se odredila najbolja alternativna metoda za predviđanje ANN modela, ukupno je obučeno 320 treniranih mreža u višeslojnom perceptronu (MLP) i ukupno 20 treniranih mreža u arhitekturi Radial Basis Function (RBF). Na temelju statistike goodness-of-fit, ANN u smislu MLP 1-9-1 uključujući dbh kao input varijablu za jednoulazna volumna predviđanja pokazao je bolju fitting sposobnost sa SSE (2.7763), Radj2 (0.9339), MSE (0.00910), RMSE (0.0954), AIC (-823.25) i SBC (-1421.81) nego onaj u ostalim proučavanim volumnim metodama koje uključuju dbh kao eksplanatornu varijablu. Za dvoulazna volumna predviđanja, što uključuju dbh i ukupnu visinu kao input varijable, ANN temeljen na MLP 2-15-1 rezultirao je boljom fitting statistikom sa SSE (0.8354), Radj2 (0.9801), MSE (0.00274), RMSE (0.0523), AIC (-579.55) and SBC (-1788.11).

CATEGORIZATION AND IDENTIFICATION OF HUMAN FACE IMAGES BY NEURAL NETWORKS: A REVIEW OF THE LINEAR AUTOASSOCIATIVE AND PRINCIPAL COMPONENT APPROACHES

1994 ◽  
Vol 02 (03) ◽  
pp. 413-429 ◽  
Author(s):  
D. VALENTIN ◽  
H. ABDI ◽  
A.J. O’TOOLE
Keyword(s):  
Neural Network ◽  
Principal Component Analysis ◽  
Face Processing ◽  
Linear Models ◽  
Network Models ◽  
Principal Component ◽  
Human Face ◽  
Neural Network Models ◽  
Face Images ◽  
Face Categorization

Recent statistical/neural network models of face processing suggest that faces can be efficiently represented in terms of the eigendecomposition of a matrix storing pixel-based descriptions of a set of face images. The studies presented here support the idea that the information useful for solving seemingly complex tasks such as face categorization or identification can be described using simple linear models (linear autoassociator or principal component analysis) in conjunction with a pixel-based coding of the faces.

scholarly journals Estimating and interpreting nonlinear receptive field of sensory neural responses with deep neural network models

eLife ◽  
2020 ◽  
Vol 9 ◽  
Author(s):  
Menoua Keshishian ◽  
Hassan Akbari ◽  
Bahar Khalighinejad ◽  
Jose L Herrero ◽  
Ashesh D Mehta ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Receptive Field ◽  
Network Models ◽  
Linear Functions ◽  
Neural Responses ◽  
Neural Network Models ◽  
Stimulus Response ◽  
Exact Function ◽  
Neurosurgical Patients

Our understanding of nonlinear stimulus transformations by neural circuits is hindered by the lack of comprehensive yet interpretable computational modeling frameworks. Here, we propose a data-driven approach based on deep neural networks to directly model arbitrarily nonlinear stimulus-response mappings. Reformulating the exact function of a trained neural network as a collection of stimulus-dependent linear functions enables a locally linear receptive field interpretation of the neural network. Predicting the neural responses recorded invasively from the auditory cortex of neurosurgical patients as they listened to speech, this approach significantly improves the prediction accuracy of auditory cortical responses, particularly in nonprimary areas. Moreover, interpreting the functions learned by neural networks uncovered three distinct types of nonlinear transformations of speech that varied considerably from primary to nonprimary auditory regions. The ability of this framework to capture arbitrary stimulus-response mappings while maintaining model interpretability leads to a better understanding of cortical processing of sensory signals.

scholarly journals Predicting Credit Rating Migration Employing Neural Network Models

2020 ◽  
pp. 65-82
Author(s):  
Michael D'Rosario ◽  
Calvin Hsieh
Keyword(s):  
Neural Network ◽  
Goodness Of Fit ◽  
Credit Rating ◽  
Predictive Accuracy ◽  
Estimation Method ◽  
Network Models ◽  
Legal Framework ◽  
Neural Network Models ◽  
Network Modelling ◽  
Extant Literature

Credit rating migration ranks amongst the most pertinent issues concerning institutional lenders and investors alike. There are a number of studies that have employed both parametric and non-parametric methodologies to forecast credit rating migration, employing machine learning methods; and notably, artificial intelligence methods becoming increasingly popular. The present study extends upon research within the extant literature employing a novel estimation method, a neural network modelling technique, herewith the MPANN (multi-layer neural network). Consistent with the extant literature, the present article identifies that the legal framework and system of taxation enacted within a polity are pertinent to predicting rating migration. However, extending upon traditional estimation techniques the study identifies that a number of different model calibrations achieve greater predictive accuracy than traditional parametric regression. Notably, the method is able to achieve superior goodness of fit and predictive accuracy in determining credit rating migration than models employed within the extant literature.

Sign in / Sign up

Export Citation Format

Share Document