Two-layer Feed-forward Neural Network Research Articles

Optical-Electrical Coordinately Modulated Memristor Based on 2D Ferroelectric RP Perovskite for Artificial Vision Applications.

Traditional artificial vision systems built using separate sensing, computing, and storage units have problems with high power consumption and latency caused by frequent data transmission between functional units. An effective approach is to transfer some memory and computing tasks to the sensor, enabling the simultaneous perception-storage-processing of light signals. Here, an optical-electrical coordinately modulated memristor is proposed, which controls the conductivity by means of polarization of the 2D ferroelectric Ruddlesden-Popper perovskite film at room temperature. The residual polarization shows no significant decay after 109-cycle polarization reversals, indicating that the device has high durability. By adjusting the pulse parameters, the device can simulate the bio-synaptic long/short-term plasticity, which enables the control of conductivity with a high linearity of ≈0.997. Based on the device, a two-layer feedforward neural network is built to recognize handwritten digits, and the recognition accuracy is as high as 97.150%. Meanwhile, building optical-electrical reserve pool system can improve 14.550% for face recognition accuracy, further demonstrating its potential for the field of neural morphological visual systems, with high density and low energy loss.

Elementary proof of Funahashi's theorem

Funahashi established that the space of two-layer feedforward neural networks is dense in the space of all continuous functions defined over compact sets in $n$-dimensional Euclidean space. The purpose of this short survey is to reexamine the proof of Theorem 1 in Funahashi \cite{Funahashi}. The Tietze extension theorem, whose proof is contained in the appendix, will be used. This paper is based on harmonic analysis, real analysis, and Fourier analysis. However, the audience in this paper is supposed to be researchers who do not specialize in these fields of mathematics. Some fundamental facts that are used in this paper without proofs will be collected after we present some notation in this paper.

Energy and economic analysis of building integrated photovoltaic thermal system: Seasonal dynamic modeling assisted with machine learning-aided method and multi-objective genetic optimization

Building integrated photovoltaic thermal (BIPV/T) systems offer a highly effective means of generating clean energy for both electricity and heating purposes in residential buildings. Hence, this article introduces a new BIPV/T system to optimally minimize the energy consumption of a household residential building. The meticulous design of the proposed BIPV/T system is accomplished through MATLAB/Simulink® dynamic modeling. Performance analysis for the BIPV/T system is performed under different seasonal conditions with in-depth techno-economic analyses to estimate the expected enhancement in the thermal, electrical, and economic performance of the system. Moreover, a sensitivity analysis is conducted to explore the impact of various factors on the energetic and economic performances of the proposed BIPV/T system. More so, the two-layer feed-forward back-propagation artificial neural network modeling is developed to accurately predict the hourly solar radiation and ambient temperature for the BIPV/T. Additionally, a multi-objective optimization using the NSGA-II method is also conducted for the minimization of the total BIPV/T plant area and maximization of the total efficiency and net thermal power of the system as well as to estimate the optimized operating conditions for input variables across different seasons within the provided ranges. The sensitivity analysis revealed that higher solar flux levels lead to increased electric output power of the BIPV/T plant, but total efficiency decreases due to higher thermal losses. Moreover, the proposed NSGA-II shows a feasible method to attain a maximum net thermal power and optimal total efficiency of 5320 W and 63% with a minimal total plant area of 32.89 m2 that attained a very low deviation index from the ideal solution. The levelised cost of electricity is obtained as 0.10 $/kWh under the optimal conditions. Thus, these findings offer valuable insights into the potential of BIPV/T systems as a sustainable and efficient energy solution for residential applications.

Optimization of Neuro-controller Application for Maximum Power Point Tracking Photovoltaic Systems Through Shannon’s Information Criteria

Due to the extremely poor efficiency of solar energy, researchers created various maximum power point tracking (MPPT) techniques with the aim of enhancing the effectiveness of photovoltaic (PV) systems. Because of their propensity to address complex problems and their non-linear characteristics, Artificial Neural Network (ANN) algorithms are the most frequently used among these MPPT techniques. Nevertheless, the performance of the ANN-based MPP tracking algorithms is contingent upon various factors, including the choice of activation function, the quantity of hidden neurons, and the training algorithm employed. Shannon’s Information Criteria (SIC) is used to determine the optimal number of hidden neurons within a single hidden layer for the neuro-controller application. In this regard, a two-layer Feed-Forward Neural Network (FFNN) was trained using MATLAB/Simulink software, incorporating varying numbers of hidden neurons. The results indicate that the two-layer FFNN with five hidden neurons has the highest performance, as demonstrated by the lowest Mean Squared Error (MSE) of 4.03 × 10-9, which is statistically significant. The successful incorporation of Gaussian noise in the simulation of an 85 kW PV system demonstrates that the ANN-based MPPT algorithm is both theoretically robust and practically viable and reliable for enhancing the efficiency of solar PV systems in real-world scenarios.

Quantitative monitoring of icing on CFRP laminate with guided wave combining forward modeling and inverse characterization

Aircraft icing is one of the critical factors for flight safety. Timely and accurate monitoring of in-flight ice accretion is essential for flight systems to take immediate and effective action to significantly improve the efficiency of subsequent de-icing processes and ensure flight safety. Since carbon fiber reinforced polymer (CFRP) are now widely used in the aviation industry, the study of ice accretion on the surfaces of CFRP with anisotropic properties is of great importance. In this work, a three-dimensional frequency domain finite element model for icing CFRP laminate is first established, via which the mechanism of mode conversion of ultrasonic guided waves (UGWs) at different ice layer thicknesses is quantitatively analyzed. Then, the time-domain finite element simulation was performed, from which the time–frequency domain spectrogram of the acquired UGW signal is obtained by short-time Fourier transform (STFT). The extracted frequency–time-of-flight (ToF) curve of the signal via STFT shows high agreement with the theoretical results, which confirms the influence of icing on UGW propagation. Finally, numerous experiments with bonded lead zirconate titanate wafers to excite and acquire UGWs are conducted with different ice layer thicknesses and lengths on CFRP. The extracted frequency–ToF curves of UGW signals are used as input to a two-layer feedforward artificial neural network (ANN) to accurately evaluate the length and thickness of the ice layer. The high reliability of this method with ANN is confirmed, which shows the potential of the proposed UGW-based method for quantitative monitoring of icing length and thickness on aircraft CFRP laminate.

Artificial neural network-based prediction model of elastic floor response spectra incorporating dynamic primary-secondary structure interaction

The evaluation of the Floor Response Spectrum (FRS) holds paramount significance in assessing the seismic behavior of secondary structures. Precise FRS prediction empowers engineers to make informed decisions concerning structural design, retrofitting, and safety precautions. This study aims to scrutinize the impact of dynamic interaction between primary and secondary structures on FRS. Both the elastic primary structure (PS) and elastic secondary structure (SS) employ a single-degree-of-freedom (SDOF) system. Governing motion equations for both coupled (with dynamic interaction) and uncoupled (without dynamic interaction) systems are formulated and solved numerically. The study investigates how variations in the vibration period of PS (Tp), tuning ratio (Tr), mass ratio (μ), and damping ratio (ξs) of SS influence FRS. The FRS impact remains minimal at μ = 0.001 (0.1%); however, with increasing mass ratio, PS-SS dynamic interaction significantly affects SS's spectral acceleration response. Coupled analysis is crucial only for secondary structures tuned to the primary structure's vibration period (0.8≤Tr≤1.2). This study utilizes two-layer feed-forward Artificial Neural Networks (ANNs) for FRS prediction. The Levenberg-Marquardt (LM) backpropagation (BP) algorithm trains the network using a comprehensive dataset. In summary, it is evident that the ANNs, once trained, enable accurate prediction of the FRS, exhibiting a R2 of 99%. Additionally, a design expression is formulated utilizing the ANN model and subsequently compared with the existing formulation.

Max-C and Min-D Projection Auto-Associative Fuzzy Morphological Memories: Theory and an Application for Face Recognition

Max-C and min-D projection auto-associative fuzzy morphological memories (max-C and min-D PAFMMs) are two-layer feedforward fuzzy morphological neural networks designed to store and retrieve finite fuzzy sets. This paper addresses the main features of these auto-associative memories: unlimited absolute storage capacity, fast retrieval of stored items, few spurious memories, and excellent tolerance to either dilative or erosive noise. Particular attention is given to the so-called Zadeh’ PAFMM, which exhibits the most significant noise tolerance among the max-C and min-D PAFMMs besides performing no floating-point arithmetic operations. Computational experiments reveal that Zadeh’s max-C PFAMM, combined with a noise masking strategy, yields a fast and robust classifier with a strong potential for face recognition tasks.

Soil moisture estimation underneath crop cover using high incidence angle C-band Sentinel-1 SAR data

SAR's return signal sensitivity to dielectric constant and penetration through vegetation makes it ideal for large-scale soil moisture studies. High incidence angle SAR data is especially useful for understanding crop characteristics. However, interpreting the return signal requires considering both crop characteristics and underlying soil moisture. In this study, an attempt has been made to assess the potential of high incidence angle C-band SAR data for soil moisture retrieval by incorporating effects of crop on Radar signal using water cloud model. The simulated VV backscatter from water cloud model by considering Leaf Area Index (LAI) and plant water content as vegetation descriptors has shown Root Mean Square Error (RMSE) of 1.28 ​dB with actual VV backscatter. Later, a two-layer feedforward neural network comprising one hidden layer with sigmoid neurons has been considered to develop retrieval models. It is observed that the ANN with input of high incidence angle C-band VV backscatter over wheat crop has soil moisture retrieval performance with correlation coefficient (R) and RMSE of 0.55 and 9.18 ​m3/m3 respectively. Another ANN model is developed to incorporate effects of crop on Radar signal by using VH backscatter, Radar Vegetation Index (RVI) and two-way attenuation factor derived from water cloud model along with VV backscatter. The overall performance of this model has been observed with R and RMSE of 0.77 and 5.81 ​m3/m3 respectively. The study results indicate that the high incidence angle SAR data may be considered for soil moisture retrieval underneath crop cover by effectively incorporating the crop effects on Radar signal. This may further extend to develop a model to extract crop biophysical parameters as well as soil moisture from Sentinel-1 SAR data.

Intelligent Prediction of Nitrous Oxide Capture in Designable Ionic Liquids

As a greenhouse gas, nitrous oxide (N2O) is increasingly damaging the atmosphere and environment, and the capture of N2O using ionic liquids (ILs) has recently attracted wide attention. Machine learning can be utilized to rapidly screen ILs suitable for N2O removal. In this study, intelligent predictions of nitrous oxide capture in designable ionic liquids are proposed based on a series of machine learning methods, including linear regression, voting, and a two-layer feed-forward neural network (TLFFNN). The voting model can utilize various algorithms and is highly generalizable to various systems. The TLFFNN model produced the most accurate prediction, with an MSE of 0.00002 and R2 of 0.9981 on test sets. The acceptable performance of the TLFFNN model demonstrates its utility as an accurate and promising candidate model for the prediction of N2O solubility in ILs over other intelligent models. Based on the analysis of the thermodynamic and molecular properties of ionic liquids, in the low-pressure zone, components of [(OH)2IM] and [AC] perform best in capturing N2O, while in the high-pressure zone, components of [(ETO)2IM] and [SCN] are best. This finding will provide new chemical insights for the industrial synthesis of ionic liquids in capturing N2O.

Neural Network Aided Homogenization Approach for Predicting Effective Thermal Conductivity of Composite Construction Materials.

Thermal conductivity is a fundamental material parameter involved in various infrastructure design guides around the world. This paper developed an innovative neural network (NN) aided homogenization approach for predicting the effective thermal conductivity of various composite construction materials. The 2-D meso-structures of dense graded asphalt mixture, porous asphalt mixture, and cement concrete were generated and divided into 2n × 2n square elements with specific thermal conductivity values. A two-layer feed-forward neural network with sigmoid hidden neurons and linear output neurons was built to predict the effective thermal conductivity of the 2 × 2 block. The Levenberg-Marquardt backpropagation algorithm was used to train the network. By repeatedly using the neural network, the effective thermal conductivities of 2-D meso-structures were calculated. The accuracy of the above NN aided homogenization approach was validated with experiment, and various factors affecting the effective thermal conductivity were analyzed. The analysis results show that the accuracy of the NN aided approach is acceptable with relative errors of 1.92~4.34% for the dense graded asphalt mixture, 1.10~6.85% for the porous asphalt mixture, and 1.13~3.14% for the cement concrete. The relative errors for all the materials are lower than 5% when the heterogeneous structures are divided into 512 × 512 elements. Ignoring the actual material meso-structures may lead to significant errors (134.01%) in predicting the effective thermal conductivity of materials with high heterogeneity such as porous asphalt mixture. While proper simplification is acceptable for dense construction composite materials. The effective thermal conductivity of composite cement-asphalt mixtures increases with higher saturation of grouted material. However, the improvement effect of the high-conductive cement paste on the composite cement-asphalt mixtures could be significantly reduced when the cement paste concentrates at the bottom of the mixture. Cracked aggregates and segregation of material components tend to decrease the effective thermal conductivity of construction materials. The NN aided homogenization approach presented in this paper is useful for selecting the effective thermal conductivity of construction materials.

Vis-NIR and SWIR hyperspectral imaging method to detect bruises in pomegranate fruit.

Fresh pomegranate fruit is susceptible to bruising, a common type of mechanical damage during harvest and at all stages of postharvest handling. Accurate and early detection of such damages in pomegranate fruit plays an important role in fruit grading. This study investigated the detection of bruises in fresh pomegranate fruit using hyperspectral imaging technique. A total of 90 sample of pomegranate fruit were divided into three groups of 30 samples, each representing purposefully induced pre-scanning bruise by dropping samples from 100 cm and 60 cm height on a metal surface. The control has no pre-scanning bruise (no drop). Two hyperspectral imaging setups were examined: visible and near infrared (400 to 1000 nm) and short wavelength infrared (1000 to 2500 nm). Region of interest (ROI) averaged reflectance spectra was implemented to reduce the image data. For all hypercubes a principal components analysis (PCA) based background removal were done prior to segmenting the region of interest (ROI) using the Evince® multi-variate analysis software 2.4.0. Then the average spectrum of the ROI of each sample was computed and transferred to the MATLAB 2022a (The MathWorks, Inc., Mass., USA) for classification. A two-layer feed-forward artificial neural network (ANN) is used for classification. The accuracy of bruise severity classification ranged from 80 to 96.7%. When samples from both bruise severity (Bruise damage induced from a 100cm and 60 cm drop heights respectively) cases were merged, class recognition accuracy were 88.9% and 74.4% for the SWIR and Vis-NIR, respectively. This study implemented the method of selecting out informative bands and disregarding the redundant ones to decreases the data size and dimension. The study developed a more compact classification model by the data dimensionality reduction method. This study demonstrated the potential of using hyperspectral imaging technology in sensing and classification of bruise severity in pomegranate fruit. This work provides the foundation to build a compact and fast multispectral imaging-based device for practical farm and packhouse applications.

A Neural Network Controller Design for the Mecanum Wheel Mobile Robot

Advanced controllers are an excellent choice for the trajectory tracking problem of Wheeled Mobile Robots (WMRs). However, these controllers pose a challenge to the hardware structure of WMRs due to the controller's complex structure and the large number of calculations needed. In that context, designing a controller with a simple structure and a small number of computations but good real-time performance is necessary in order to improve the tracking accuracy for the WMRs without requiring high hardware architecture. In this work, a neural network controller with a simple structure for the trajectory-tracking of a Mecanum-Wheel Mobile robot (MWMR) based on a reference controller is proposed. A two-layer feedforward neural network is designed as a tracking controller for the robot. The neural network is trained with a sample input-output data set so that the error between the neural network output and the reference control signal of the supervisory controller is minimal. The neural network parameters are trained to update over time. The simulation results verified the effectiveness of the neural network controller, whose parameters are tuned adaptively to ensure a fast convergence to the desired Bézier trajectory.

Learning knot invariants across dimensions

We use deep neural networks to machine learn correlations between knot invariants in various dimensions. The three-dimensional invariant of interest is the Jones polynomial J(q)J(q), and the four-dimensional invariants are the Khovanov polynomial \text{Kh}(q,t)Kh(q,t), smooth slice genus gg, and Rasmussen’s ss-invariant. We find that a two-layer feed-forward neural network can predict ss from \text{Kh}(q,-q^{-4})Kh(q,−q−4) with greater than 99%99% accuracy. A theoretical explanation for this performance exists in knot theory via the now disproven knight move conjecture, which is obeyed by all knots in our dataset. More surprisingly, we find similar performance for the prediction of ss from \text{Kh}(q,-q^{-2})Kh(q,−q−2), which suggests a novel relationship between the Khovanov and Lee homology theories of a knot. The network predicts gg from \text{Kh}(q,t)Kh(q,t) with similarly high accuracy, and we discuss the extent to which the machine is learning ss as opposed to gg, since there is a general inequality |s| ≤2g|s|≤2g. The Jones polynomial, as a three-dimensional invariant, is not obviously related to ss or gg, but the network achieves greater than 95%95% accuracy in predicting either from J(q)J(q). Moreover, similar accuracy can be achieved by evaluating J(q)J(q) at roots of unity. This suggests a relationship with SU(2)SU(2) Chern—Simons theory, and we review the gauge theory construction of Khovanov homology which may be relevant for explaining the network’s performance.

Hybrid encryption technique: Integrating the neural network with distortion techniques.

This paper proposes a hybrid technique for data security. The computational model of the technique is grounded on both the non-linearity of neural network manipulations and the effective distortion operations. To accomplish this, a two-layer feedforward neural network is trained for each plaintext block. The first layer encodes the symbols of the input block, making the resulting ciphertext highly uncorrelated with the input block. The second layer reverses the impact of the first layer by generating weights that are used to restore the original plaintext block from the ciphered one. The distortion stage imposes further confusion on the ciphertext by applying a set of distortion and substitution operations whose functionality is fully controlled by random numbers generated by a key-based random number generator. This hybridization between these two stages (neural network stage and distortion stage) yields a very elusive technique that produces ciphertext with the maximum confusion. Furthermore, the proposed technique goes a step further by embedding a recurrent neural network that works in parallel with the first layer of the neural network to generate a digital signature for each input block. This signature is used to maintain the integrity of the block. The proposed method, therefore, not only ensures the confidentiality of the information but also equally maintains its integrity. The effectiveness of the proposed technique is proven through a set of rigorous randomness testing.

Prognosis of LED lumen degradation using Bayesian optimized neural network approach

Light Emitting Diodes (LEDs) are among the most widely used electronic devices for everyday lighting applications due to their durability over incandescent lamps. However, there are no standardized approaches to predict the reliability of LEDs as they are manufactured and tested as per the user requirements and applications. This dramatically limits developing generic prognostic algorithms pertaining to predicting the remaining useful life (RUL) of LEDs. In this study, we propose a Bayesian optimized neural network approach to predict the lumen degradation trends of LEDs. The proposed method does not require an accurate physical model representing the LED degradation behavior and does not require a large amount of degradation data. We have used a particle filter algorithm to train a simple two-layer feedforward neural network model and use the trained model to predict the lumen degradation of LEDs. Also, the weight decay issues commonly encountered in particle filter algorithm are addressed using three different resampling strategies and particle roughening method. To evaluate the effectiveness of the proposed approach, Root Mean Squared Error (RMSE) and Relative Accuracy (RA) were used as the prognostic metrics.

A Comparative Study on Prediction of Cutting Force using Artificial Neural Network and Genetic Algorithm during Machining of Ti-6Al-4V

The purpose of this comparative study is to improve the predictive accuracy of the cutting force during the turning of Ti-6Al-4V on a lathe machine. By optimizing the machining process parameters such as cutting speed, feed rate, and depth of cut, the cutting force in the machining process can be improved significantly. Cutting force is one of the crucial characteristics that must be monitored during the cutting process in order to enhance tool life and the surface finish of the workpiece. This paper is based on the experimental dataset of cutting forces collected during the turning of titanium alloy under the Minimum Quantity Lubrication (MQL) condition. To predict the cutting forces, two machine learning techniques are explored. Firstly, a black-box model called an Artificial Neural Network (ANN) is proposed to predict cutting force. Using the Levenberg-Marquardt algorithm, a two-layered feedforward neural network is built in MATLAB to predict cutting force. The second model to be implemented was the Genetic Algorithm (GA), a white-box model. GA is an optimization technique which is based on Darwinian theories. It is a probabilistic method of searching, unlike most other search algorithms, which require definite inputs. Using symbolic regression in HeuristicLab, a GA model is developed to estimate cutting force. The anticipated values of cutting forces for both models were compared. Since the ANN model had fewer errors, it was ascertained that the particular model is preferable for machining process optimization.

Multi-Dimensional Feature Combination Method for Continuous Blood Pressure Measurement Based on Wrist PPG Sensor.

The cuff-less blood pressure (BP) monitoring method based on photoplethysmo- gram (PPG) makes it possible for long-term BP monitoring to prevent and treat cardiovascular and cerebrovascular events. In this paper, a portable BP prediction system based on feature combination and artificial neural network (ANN) is implemented. The robustness of the model is improved from three aspects. Firstly, an adaptive peak extraction algorithm was used to improve the accuracy of peaks and troughs detection. Secondly, multi-dimensional features were extracted and fused, including three groups of PPG-based features and one group of demographics-based features. Finally, a two-layer feedforward artificial neural networks algorithm was used for regression. Thirty-three subjects distributed in the three BP groups were recruited. The proposed method passed the European Society of Hypertension International Protocol revision 2010 (ESP-IP2). Experimental results show that the proposed method exhibits good accuracy for a diverse population with an estimation error of -0.07 ± 4.47 mmHg for SBP and 0.00 ± 3.61 mmHg for DBP. Moreover, the model tracked the BP of two subjects for half a month, laying the foundation work for daily BP monitoring. This work will contribute to the long-term wellness management and rehabilitation process, enabling timely detection and improvement of the user's physical health.

Surface roughness prediction applying artificial neural network at micro machining

Micro machining forms an integral part of modern manufacturing technology. In order to satisfy the most important demands, like cost reduction, quality improvement, and production efficiency, this process has to be implemented into the framework of Industry 4.0, too. Defect reduction of products can be realized by beneficial trends of Industry 4.0, such as application of sensors driven AI based predictive models. Surface roughness is one of the main indicators, which can be used for evaluation of the machined surface. The aim of this research is to develop a predictive surface roughness model in the case of micro milling of hard materials. Most important scientific knowledge related to micro milling, and application possibilities of AI based predictive models in the case of chip removal processes were summarised. Micro milling experiments on hot work tool steel (AISI H13) were carried out in order to determine the effect of different cutting parameters on surface roughness. The effects of the cutting parameters were investigated by the analysis of variance (ANOVA). Two-layer feedforward artificial neural network (ANN) based model was trained, validated and tested by means of experimental results. It was found that the trained model has an average MSE, RMSE and R values of 0.00063850, 0.019116835 and 0.9948, while the additional validation test provided 0.00038148, 0.019531513 and 0.8545, respectively, which are statistically significant. The developed model can effectively predict the surface roughness under the investigated circumstances.

Standardizing continuous data classifications in a virtual T-maze using two-layer feedforward networks

There continues to be difficulties when it comes to replication of studies in the field of Psychology. In part, this may be caused by insufficiently standardized analysis methods that may be subject to state dependent variations in performance. In this work, we show how to easily adapt the two-layer feedforward neural network architecture provided by Huang1 to a behavioral classification problem as well as a physiological classification problem which would not be solvable in a standardized way using classical regression or “simple rule” approaches. In addition, we provide an example for a new research paradigm along with this standardized analysis method. This paradigm as well as the analysis method can be adjusted to any necessary modification or applied to other paradigms or research questions. Hence, we wanted to show that two-layer feedforward neural networks can be used to increase standardization as well as replicability and illustrate this with examples based on a virtual T-maze paradigm2–5 including free virtual movement via joystick and advanced physiological data signal processing.

Resource optimization of MIMO using neural network for 5G communication

The principal objective of this work is to optimize the resources of massive MIMO techniques by an application of a neural network algorithm in order to reduce the manufacturing cost of 5G technology within the wireless peripheries of tropical regions. For plenty of annual precipitation in the tropical regions, the radio signals of these areas get attenuated in communicating any message i.e. signal fading occurs. Various diversity techniques have been used for diminishing signal fading in these tropical areas. MIMO technique is a spatial diversity technique used to minimize signal fading. In this paper, 1 × 1 to 256 × 256 MIMO techniques have been designed with the help of MATLAB and Xilinx system generator, and an 8 × 8 MIMO has been implemented in Xilinx Kintex7 FPGA. A CMOS logic technique has been incorporated to optimize the resources of different MIMO techniques, and for this purpose a two-layer feed forward neural network algorithm has been adopted. A comparison of the theoretical values obtained from the design by using both Xilinx system generator and Simulink with the values derived from the design by using CMOS logic gates, has been displayed in this paper. For neural network optimization, three types of training algorithms have been applied.