An Extreme Learning Machine (ELM) Predictor for Electric Arc Furnaces v-i Characteristics

Size: px
Start display at page:

Download "An Extreme Learning Machine (ELM) Predictor for Electric Arc Furnaces v-i Characteristics"

Transcription

1 An Extreme Learning Machine (ELM) Predictor for Electric Arc Furnaces v-i Characteristics Salam Ismaeel, Ali Miri, Alireza Sadeghian and Dharmendra Chourishi Department Computer Science, Ryerson University, Toronto, Canada salam.ismaeel, ali.miri, asadeghi, Abstract This paper presents an Extreme Learning Machine (ELM) time series prediction strategy to estimate the current and voltage behaviour of an Electric Arc Furnace (EAF). The proposed ELM predictor is designed for both long and short term predictions of the v-i characteristics of an EAF. The proposed predictor is evaluated using two real sensors outputs collected over different time periods with a rate of 2000 samples per second, and its performance is compared against Feed-Forward Neural Networks (FFNN), Radial Basis Functions (RBF) and Adaptive Neuro-Fuzzy Inference Systems (ANFIS) algorithms. Experimental results obtained show the proposed ELM predictor to have superior speed and stability behaviour, while obtaining similar error values to comparable techniques. Index Terms Extreme Learning Machine (ELM), neural networks, real-time prediction, Electric Arc Furnace (EAF) I. INTRODUCTION In today s highly inter-connected world, large volumes of data are continuously created from various sources such as sensors, actuators, and data processors. These types of devices and resources enable real-time sensing, capturing, collection of many connected systems. While the volume and the velocity characteristics of data - collected, stored, and processed - pose many computational challenges [1, 2], the use of such data provide an invaluable tool in many different settings, including environmental monitoring, and industrial, business and human-centric pervasive applications. This paper focuses on the modelling of large amounts of sensor data collected from Electric Arc Furnaces (EAFs). EAFs are widely used in the steelmaking industry [3], and they are considered to be one of the highest electrical energy consumers on power grids. The rising cost of energy has put pressure on the steel industry to improve their process control systems to conserve energy without sacrificing quality and equipments [4]. An accurate and reliable estimate of the v-i characteristics of an EAF can provide an effective tool in the energy management of such devices. The measured v-i characteristics are nonlinear, discrete, and time-variant, and have non-stationary stochastic behaviour, with dependent variables. EAFs need significant involvement of human operators with expert operational knowledge in very noisy environments [3, 4, 5]. High speed sensors are also essential to keep up with such highly dynamic environments. However, conventional techniques are typically incapable of handling storage or processing of information from such sensors [6], due to the exponential growth of volume, variety and velocity of data generated. Meanwhile, machine learning techniques together with advances in available computational power have come to play a vital role in big data analytics and knowledge discovery [7]. The development of fast and efficient algorithms for real-time processing and prediction of the data has gained attention in recent years in both scientific research and engineering applications. Several machine learning methods have been used to predict time series data [8]. One of the most powerful tools which has attracted considerable attention is the Extreme Learning Machine (ELM), which has extremely fast learning speeds and good generalization performance [9, 10]. In this work, a simplified ELM architecture is proposed to estimate the next state of the voltage v and current i in an EAF described in [11]. This prediction only depends on outputs of two high-speed sensors that are used to provide real-time monitoring of an EAF s voltage and current. The rest of the paper is organized as follows. Section II provides an introduction of the ELM algorithm. In Section III, we describe the proposed ELM algorithm for time-series prediction. Section IV includes the comparison results of the proposed prediction system with other systems in the literature, followed by the conclusions in Section V. II. EXTREME LEARNING MACHINES ELMs are Single, hidden Layer Feed-forward neural Networks (SLFN), consisting of one input layer, one hidden layer and one output layer, as shown in Figure 1 [10]. In this architecture, n is the number of inputs; l denotes the number of hidden neurons; m is the number of output neurons; w ji denotes the weight between the i th neuron in the input layer and j th neuron in the hidden layer; β jr denotes the weights between the j th neuron in the hidden layer and r th neuron in the output layer; and b j is the threshold in the j th hidden layer. Let X(k) and Y (k) be the input and output vectors at sample k, respectively. Also, let W be the input weight matrix (l n), B be the output weight matrix (l m), and b be the bias vector. That is, X(k) = [x 1 (k) x 2 (k)... x n (k)] T (1) Y (k) = [y 1 (k) y 2 (k)... y m (k)] T (2)

2 and the output of the j th neuron is: l y j = β ij g(w i x j + b i ) (9) If target vector T is given by: t 1 t 2 T =. (10) Figure 1: The ELM architecture w 11 w 12 w 1n w 21 w 22 w 2n W = w l1 w l2 w ln β 11 β 12 β 1m β 21 β 22 β 2m B = β l1 β l2 β lm b 1 b 2 b =. b l l 1 l n l m Suppose there are Q data samples, yeilding input matrix X and output matrix Y: x 11 x 12 x 1Q x 21 x 22 x 2Q X = (6) x n1 x n2 x nq n Q y 11 y 12 y 1Q y 21 y 22 y 2Q Y = (7) y m1 y m2 y mq m Q For a sigmoid activation function, the output of the i th hidden neuron will be: (3) (4) (5) g i (wx + bi) (8) where w i = [w i1, w i2,..., w in ] and x j = [x j1, x j2,..., x nj ] T t Q m Q where t j = [t 1j,..., t mj ] m 1. Then using Equation 9, target matrix T can be written as: l β i1 g i (w i x j + bi) l β i2 g i (w i x j + bi) T = (11).. l β im g i (w i x j + bi) By separating β and the function g in Equation 11 we get: Hβ = T (12) where β = [β 1, β 2,..., β l ] T and H is the activation matrix given by: H(w 1,..., w l, b 1,..., b l, x 1,..., x Q ) = g(w 1 x 1 + b 1 ) g(w 2 x 1 + b 2 ) g(w l x 1 + b l ) g(w 1 x 2 + b 1 ) g(w 2 x 2 + b 2 ) g(w l x 2 + b l ).... (13) g(w 1 x Q + b 1 ) g(w 2 x Q + b 2 ) g(w l x Q + b l ) H is called the hidden-layer output matrix of the network. The i th column of H is the i th hidden node s output vector with respect to inputs x 1, x 2,, x n and the j th row of H is the output vector of the hidden layer with respect to input x j [12]. Huang et al [13, 14] showed that the input weights w j and biases b j for j = 1, 2,, l associated with the hidden layer can be assigned random values and require no additional training. Also, the output weights β i, i = 1, 2,, m are determined through learning from the training instances by solving the following objective function: min }{{} Hβ T (14) β and its solution is [15]: ( ) 1 I ˆβ = R + HT H H T T = GH T T (15) for l < Q and ˆβ = H T ( I R + HT H) 1 T = H T GT (16)

3 for l > Q, where I is identity matrix, and C is a regularization parameter. G is given by: ( ) 1 I G = C + HT H (17) The output for new input data (Ŷ (k)) is calculate as: Ŷ (k) = h ˆβ (18) where h is H matrix for the new input. Based on the ELM set of equations, one can note the following: i) The input weights and bias W and b can be randomly generated and used without tuning. ii) The output weights β can be tuned using simple, fast operations. iii) The only two parameters which a user needs to select are: the number of hidden neurons (l) and the regularization parameter (C) (We will discuss these parameters in more detail in the next section). The main advantages of ELM over conventional gradient based learning methods are [8, 16, 17]: It avoids many problems such as stopping criteria, learning rate, learning epochs, and local minima cannot be accomodated by classical methods. It is a fast learning algorithm, because the learning of ELM is a one-pass algorithm without re-iteration. It can obtain better generalized performance than the Back Propagation (BP) algorithm in most cases. It is suitable for almost all nonlinear activation functions. Some recent work has proposed major modifications to the original ELM architecture used in prediction applications. Wang and Han [18] suggested a Multiple Kernel Extreme Learning Machine (MKELM) for multivariate time series predictions. The proposed MKELM was then used to predict the Lorenz chaotic time series. Bian et al. [19] suggested a combination of the Logistic Regression (LR) model and the ELM algorithm in order to improve the real-time performance of a flare forecasting method. The LR model is used to map three parameters of each active region onto four probabilities. The ELM is then used to map the four probabilities into a binary label which represents the final output. The LR model not only improves the learning speed of prediction, but also the accuracy of the flare prediction. Wang and Han [8] improved the ELM performance by adding the Levenberg-Marquardt algorithm, in which a Hessian matrix and gradient vectors are calculated iteratively, to implement an n-line sequential prediction. Their modification improved the simulation results in both an artificial and real-world multivariate time series. Other modifications made to the original ELM algorithm, developed by Huang in 2004 [13] wich have focused on improving some of the algorithm s shortcomings or modifying its performance include Regularized ELM (RELM) and Extreme Learning Machine with Kernel (KELM) [20], Self-adaptive extreme learning machine [17], Multilayer Figure 2: A sample of v-i waveforms extreme learning machine (MLELM) [15], and Deep-Structure ELM (DS-ELM) [16]. The main contribution of this work is to use the basic structure of the ELM to configure the proposed predictor for v-i characteristics of EAR. This structure consists of only 4 neuron inputs, 10 hidden and one output neuron. III. PROPOSED ELM PREDICTOR Figure 2 shows a sample of EAF voltage and current waveforms. It can be observed that EAF imbalances create even and odd harmonics, inter-harmonics and footage flicker. Voltage and current are highly dynamic during its operation and this dynamic behaviour makes it challenging to model the associated characteristic accurately [11]. The proposed ELM predictor can be used either to predict the voltage v(k + 1), or the current i(k + 1), based on the existing and previous state of the voltage and the current. For example, in all experiments presented in this paper, we use v(k), v(k 1), i(k) and i(k 1) to predict the value of v(k+1). Using the proposed ELM prediction architecture in Figure 3, one obtains a model with 4 inputs and one output. Algorithm 1 describes the steps of the of the proposed ELM predictor. The inputs and output are scaled to [ 1 1]. As shown in the next section, different values will be used for the two selected parameters C and l and we will show that this will affect the performance of the network. In our experiments, the best values for C and l were found to be 5000 and 10, respectively. The most expensive operation in constructing the model is the calculation of the inverse matrix term in Equation 15. For large dimensions, one can use specialized inverse calculation algorithms, such as generalized Moore-Penrose which has been used by Samet and Miri [21] to increase the speed of computation. IV. EXPERIMENTAL RESULTS Four data sets were used to evaluate the proposed predictor. Each set consisted of 5000 samples representing measurements

4 Figure 4: MSE vs sampling rate Figure 3: Proposed ELM architecture Algrithm1: Proposed ELM Algorithm 1: Normalize the input data sets v(k) and i(k), scaled to [ 1 1] 2: Normalize the required output v(k + 1) and/or i(k + 1), scaled to [ 1 1] 3: Choose the values of the regulation parameter C and the number of hidden neurons l 4: Randomly initialize the weights w i and bias values b i 5: Calculate the hidden layer output matrix H(X, w, β) using Equation 13, where X(k) = [v(k) v(k 1) i(k) i(k 1)] T 6: Calculate the inverse matrix term G in Equation 17 7: Calculate the output of the hidden layer using the new unknown input h(z, w, β) where z = v(k + 1) 8: Calculate the predicted output, of the ELM ˆv(k +1) using Equation 18: ˆv(k + 1) = hgh T T (19) of 2.5 second, where the sampling frequency was 1920 Hz. We used 80% of the available data, that is (4000) samples of each set to update the network parameters and the last 20%, that is 1000 samples, to test the behaviour of the networks. Figure 4 gives the relation between step size, varied from 1 to 50, and prediction mean square error (MSE) for the four data sets. Given that no big difference in MSE was observed, we have multiplied the sampling rate by 10 for all networks used in order to reduce the number of required calculations in the training phase. In other words, 400 samples were used to represent 4000 samples from the real data. After adaptations, all networks are then used to predict the next 1000 samples. In order to illustrate the effectiveness of the ELM predictor, the proposed algorithm is compared with Feed Forward Neural Networks (FFNNs) using the Levenberg-Marquardt learning rule [22], Radial Basis Functions (RBFs) [19, 23] and Adaptive Neuro-Fuzzy Inference Systems (ANFISs) [24, 25, 26, 27]. Figure 5: MSE vs number of hidden neuron l The activation functions for FFNN and ELM are chosen as the Tan-Sigmoid function; while the Gaussian functions are used as the activation functions for the RBF and ANFIS. The input weights are randomly generated from the uniform interval [ 1, 1]. The inputs and output are normalized to [ 1, 1]. Figure 5 shows the relation between the number of hidden neurons, l and the MSE for the the proposed predictor. The relation between the regulation parameter C and the MSE is given in Figure 6 for the four different data sets used in this study. As we see from Figure 5 and Figure 6, l = 10 and C = 5000 are the best values for the proposed predictor with the given data set. To compare the behaviour of the ELM predictor with other algorithms, we will keep the number of hidden neurons constant (l = 10) for both FFNN and RBF, while keeping the minimum error required during the learning phase for both sets to , with 1000 epochs as maximum iterations. The parameter settings of the ANFIS selected after several tests to overcome the over-fitting problem with minimum error were: number of nodes = 55, total number of parameters is 104 (24 nonlinear and 80 linear), and number of fuzzy rules = 16. Figure 7 shows the output of the four networks compared to the desired output. In terms of accuracy, all algorithms gave a good estimate for the specified data. However, the

5 Table I: Algorithms performance Data FFNN RBF ANFIS ELM Set1 Speed MSE Set2 Speed MSE Set3 Speed MSE Set4 Speed MSE an EAF. In our proposed model, the ELM technique can use all previous knowledge to predict the next sensor output, while overcoming the learning time problem. Our proposed ELM predictor performance compares well with other machine learning techniques, with high-speed performance while producing minimal errors. Figure 6: MSE vs regulation parameter C Figure 7: Desired and predicted outputs of data set1 main feature of the proposed ELM predictor is that it does not need a learning phase. This implies that we can calculate the predicted values of every sample from the previous 4000 samples. In other words, we can always calculate the parameter of the ELM predictor, whereas in other algorithm, the training phase must always be done first, before predicting new values, making these algorithm less suitable for applications such as EAFs. This is because each EAF is unique in randomness of the arc current, the grade of steel, the size of furnace and the type and quality of scrap [28]. In other words, the selection of a fixed model for the predictor is impossible. Table I gives a comparison between all algorithms in speed in seconds and mean square error (MSE) for four different sets of 5000 samples for the same sensor. From the simulation results, we notice that the performance of the ELM predictor is quite good compared to the other algorithms. Although the ANFIS or RBF sometimes more accurate, the ELM predictor is faster than all algorithms, which is demanded by high-speed applications. V. CONCLUSION This work explored the application of ELM in prediction of EAFs devices. An ELM predictor was designed and implemented using four real v-i characteristic data sets of REFERENCES [1] B. K. Tannahill and M. Jamshidi, Big data analytic paradigms-from pca to deep learning, in Proceedings of the 2014 The Association for the Advancement of Artificial Intelligence Symposium (AAAI 2014), 2014, pp [2] A. Zaslavsky, C. Perera, and D. Georgakopoulos, Sensing as a service and big data, arxiv preprint arxiv: , [3] L. Livi, E. Maiorino, A. Rizzi, and A. Sadeghian, On the long-term correlations and multifractal properties of electric arc furnace time series, arxiv preprint arxiv: , [4] F. Janabi-Sharifi and G. Jorjani, An adaptive system for modelling and simulation of electrical arc furnaces, Control Engineering Practice, vol. 17, no. 10, pp , [5] G. Chang, M.-F. Shih, Y.-Y. Chen, and Y.-J. Liang, A hybrid wavelet transform and neural-network-based approach for modelling dynamic voltage-current characteristics of electric arc furnace, IEEE Transactions on Power Delivery, vol. 29, no. 2, pp , April [6] S. Suthaharan, Big data classification: Problems and challenges in network intrusion prediction with machine learning, ACM SIGMETRICS Performance Evaluation Review, vol. 41, no. 4, pp , [7] X. wen Chen and X. Lin, Big data deep learning: Challenges and perspectives, IEEE Access, vol. 2, pp , [8] X. Wang and M. Han, Improved extreme learning machine for multivariate time series online sequential prediction, Engineering Applications of Artificial Intelligence, vol. 40, no. 0, pp , [9] X. Mo, Y. Wang, and X. Wu, Hypoglycemia prediction using Extreme Learning Machine (ELM) and regularized ELM, in Proceedings of The 25th Chinese Control and Decision Conference (CCDC 2013), May 2013, pp

6 [10] S. Ismaeel, A. Miri, and D. Chourishi, Using the extreme learning machine (elm) technique for heart disease diagnosis, in In Proceedings of The 2015 IEEE Canada International Humanitarian Technology Conference (IHTC2015), May 2015, pp [11] A. Sadeghian and J. Lavers, Dynamic reconstruction of nonlinear characteristic in electric arc furnaces using adaptive neuro-fuzzy rule-based networks, Applied Soft Computing, vol. 11, no. 1, pp , [12] G. Feng, G.-B. Huang, Q. Lin, and R. Gay, Error minimized extreme learning machine with growth of hidden nodes and incremental learning, IEEE Transactions on Neural Networks, vol. 20, no. 8, pp , [13] G.-B. Huang, Q.-Y. Zhu, and C.-K. Siew, Extreme learning machine: Theory and applications, Neurocomputing, vol. 70, no. 1-3, pp , 2006, in Proceeding of The Neural Networks Selected Papers from the 7th Brazilian Symposium on Neural Networks (SBRN 04) 7th Brazilian Symposium on Neural Networks. [14] G.-B. Huang, H. Zhou, X. Ding, and R. Zhang, Extreme Learning Machine for regression and multiclass classification, IEEE Transactions onsystems, Man, and Cybernetics, Part B: Cybernetics, vol. 42, no. 2, pp , April [15] S. Ding, N. Zhang, X. Xu, L. Guo, and J. Zhang, Deep Extreme Learning Machine and its application in eeg classification, Mathematical Problems in Engineering, [16] M. M. Ma and B. He, Study on deep structure of Extreme Learning Machine (ds-elm) for datasets with noise, in Advanced Materials Research, vol Trans Tech Publ, 2014, pp [17] G.-G. Wang, M. Lu, Y.-Q. Dong, and X.-J. Zhao, Self-adaptive Extreme Learning Machine, Neural Computing and Applications, pp. 1 13, [18] X. Wang and M. Han, Multivariate time series prediction based on multiple kernel extreme learning machine, in Proceedings of 2014 International Joint Conference on The Neural Networks (IJCNN 2014), July 2014, pp [19] R. Kruse, C. Borgelt, F. Klawonn, C. Moewes, M. Steinbrecher, and P. Held, Radial basis function networks, in Computational Intelligence, ser. Texts in Computer Science. Springer London, 2013, pp [20] G.-B. Huang, An insight into Extreme Learning Machines: Random neurons, random features and kernels, Cognitive Computation, vol. 6, no. 3, pp , [21] S. Samet and A. Miri, Privacy-preserving back-propagation and extreme learning machine algorithms, Data and Knowledge Engineering, vol. 79,80, no. 0, pp , [22] D. Howard, M. Beale, and M. Hagan, Neural network toolbox 6. User s guide. The MathWorks Inc., [23] A. Sadeghian and J. Lavers, Application of radial basis function networks to model electric arc furnaces, in In Proceedings of The International Joint Conference on Neural Networks (IJCNN 99), vol. 6. IEEE, 1999, pp [24], Neuro-fuzzy predictors for the approximate prediction of v-i characteristic of electric arc furnaces, in Proceedings of The 19th International Conference of the North American Fuzzy Information Processing Society (NAFIPS 2000), 2000, pp [25], Application of adaptive fuzzy logic systems to model electric arc furnaces, in In Proceedings of The.18th International Conference of the North American Fuzzy Information Processing Society (NAFIPS 99). IEEE, 1999, pp [26] A. F. Lutfy, A. M. Hassan, and S. A. Ismaeel, Error estimation for an integrated gps/ins system using adaptive neuro-fuzzy technique, Iraqi Journal of Computers, Communications, Control and Systems Engineering, vol. 9, no. 1, pp , [27] Mathworks, Fuzzy Logic Toolbox R2015a. User s guide. The MathWorks, Inc., [28] J. J. Snell, Improved modeling and optimal control of an electric arc furnace, Master Thesis, University of Iowa, July 2010.

Electric Load Forecasting Using Wavelet Transform and Extreme Learning Machine

Electric Load Forecasting Using Wavelet Transform and Extreme Learning Machine Electric Load Forecasting Using Wavelet Transform and Extreme Learning Machine Song Li 1, Peng Wang 1 and Lalit Goel 1 1 School of Electrical and Electronic Engineering Nanyang Technological University

More information

Bidirectional Representation and Backpropagation Learning

Bidirectional Representation and Backpropagation Learning Int'l Conf on Advances in Big Data Analytics ABDA'6 3 Bidirectional Representation and Bacpropagation Learning Olaoluwa Adigun and Bart Koso Department of Electrical Engineering Signal and Image Processing

More information

Bearing fault diagnosis based on EMD-KPCA and ELM

Bearing fault diagnosis based on EMD-KPCA and ELM Bearing fault diagnosis based on EMD-KPCA and ELM Zihan Chen, Hang Yuan 2 School of Reliability and Systems Engineering, Beihang University, Beijing 9, China Science and Technology on Reliability & Environmental

More information

A Wavelet Neural Network Forecasting Model Based On ARIMA

A Wavelet Neural Network Forecasting Model Based On ARIMA A Wavelet Neural Network Forecasting Model Based On ARIMA Wang Bin*, Hao Wen-ning, Chen Gang, He Deng-chao, Feng Bo PLA University of Science &Technology Nanjing 210007, China e-mail:lgdwangbin@163.com

More information

A Novel Activity Detection Method

A Novel Activity Detection Method A Novel Activity Detection Method Gismy George P.G. Student, Department of ECE, Ilahia College of,muvattupuzha, Kerala, India ABSTRACT: This paper presents an approach for activity state recognition of

More information

A BAYESIAN APPROACH FOR EXTREME LEARNING MACHINE-BASED SUBSPACE LEARNING. Alexandros Iosifidis and Moncef Gabbouj

A BAYESIAN APPROACH FOR EXTREME LEARNING MACHINE-BASED SUBSPACE LEARNING. Alexandros Iosifidis and Moncef Gabbouj A BAYESIAN APPROACH FOR EXTREME LEARNING MACHINE-BASED SUBSPACE LEARNING Alexandros Iosifidis and Moncef Gabbouj Department of Signal Processing, Tampere University of Technology, Finland {alexandros.iosifidis,moncef.gabbouj}@tut.fi

More information

Lecture 5: Logistic Regression. Neural Networks

Lecture 5: Logistic Regression. Neural Networks Lecture 5: Logistic Regression. Neural Networks Logistic regression Comparison with generative models Feed-forward neural networks Backpropagation Tricks for training neural networks COMP-652, Lecture

More information

ARTIFICIAL NEURAL NETWORK WITH HYBRID TAGUCHI-GENETIC ALGORITHM FOR NONLINEAR MIMO MODEL OF MACHINING PROCESSES

ARTIFICIAL NEURAL NETWORK WITH HYBRID TAGUCHI-GENETIC ALGORITHM FOR NONLINEAR MIMO MODEL OF MACHINING PROCESSES International Journal of Innovative Computing, Information and Control ICIC International c 2013 ISSN 1349-4198 Volume 9, Number 4, April 2013 pp. 1455 1475 ARTIFICIAL NEURAL NETWORK WITH HYBRID TAGUCHI-GENETIC

More information

Modelling of Pehlivan-Uyaroglu_2010 Chaotic System via Feed Forward Neural Network and Recurrent Neural Networks

Modelling of Pehlivan-Uyaroglu_2010 Chaotic System via Feed Forward Neural Network and Recurrent Neural Networks Modelling of Pehlivan-Uyaroglu_2010 Chaotic System via Feed Forward Neural Network and Recurrent Neural Networks 1 Murat ALÇIN, 2 İhsan PEHLİVAN and 3 İsmail KOYUNCU 1 Department of Electric -Energy, Porsuk

More information

Machine Learning for Large-Scale Data Analysis and Decision Making A. Neural Networks Week #6

Machine Learning for Large-Scale Data Analysis and Decision Making A. Neural Networks Week #6 Machine Learning for Large-Scale Data Analysis and Decision Making 80-629-17A Neural Networks Week #6 Today Neural Networks A. Modeling B. Fitting C. Deep neural networks Today s material is (adapted)

More information

This is the published version.

This is the published version. Li, A.J., Khoo, S.Y., Wang, Y. and Lyamin, A.V. 2014, Application of neural network to rock slope stability assessments. In Hicks, Michael A., Brinkgreve, Ronald B.J.. and Rohe, Alexander. (eds), Numerical

More information

Neural Networks and Deep Learning

Neural Networks and Deep Learning Neural Networks and Deep Learning Professor Ameet Talwalkar November 12, 2015 Professor Ameet Talwalkar Neural Networks and Deep Learning November 12, 2015 1 / 16 Outline 1 Review of last lecture AdaBoost

More information

Deep Feedforward Networks

Deep Feedforward Networks Deep Feedforward Networks Liu Yang March 30, 2017 Liu Yang Short title March 30, 2017 1 / 24 Overview 1 Background A general introduction Example 2 Gradient based learning Cost functions Output Units 3

More information

ECE521 Lectures 9 Fully Connected Neural Networks

ECE521 Lectures 9 Fully Connected Neural Networks ECE521 Lectures 9 Fully Connected Neural Networks Outline Multi-class classification Learning multi-layer neural networks 2 Measuring distance in probability space We learnt that the squared L2 distance

More information

Feed-forward Network Functions

Feed-forward Network Functions Feed-forward Network Functions Sargur Srihari Topics 1. Extension of linear models 2. Feed-forward Network Functions 3. Weight-space symmetries 2 Recap of Linear Models Linear Models for Regression, Classification

More information

Olfactory Target/Background Odor Detection via Self-expression Model

Olfactory Target/Background Odor Detection via Self-expression Model Olfactory Target/Background Odor Detection via Self-expression Model Pingling Deng College of Communication Engineering Chongqing University Chongqing 400044, China dengpl@cqu.edu.cn Lei Zhang College

More information

Data Mining Part 5. Prediction

Data Mining Part 5. Prediction Data Mining Part 5. Prediction 5.5. Spring 2010 Instructor: Dr. Masoud Yaghini Outline How the Brain Works Artificial Neural Networks Simple Computing Elements Feed-Forward Networks Perceptrons (Single-layer,

More information

CSC242: Intro to AI. Lecture 21

CSC242: Intro to AI. Lecture 21 CSC242: Intro to AI Lecture 21 Administrivia Project 4 (homeworks 18 & 19) due Mon Apr 16 11:59PM Posters Apr 24 and 26 You need an idea! You need to present it nicely on 2-wide by 4-high landscape pages

More information

Computational statistics

Computational statistics Computational statistics Lecture 3: Neural networks Thierry Denœux 5 March, 2016 Neural networks A class of learning methods that was developed separately in different fields statistics and artificial

More information

SARIMA-ELM Hybrid Model for Forecasting Tourist in Nepal

SARIMA-ELM Hybrid Model for Forecasting Tourist in Nepal Volume-03 Issue-07 July-2018 ISSN: 2455-3085 (Online) www.rrjournals.com [UGC Listed Journal] SARIMA-ELM Hybrid Model for Forecasting Tourist in Nepal *1 Kadek Jemmy Waciko & 2 Ismail B *1 Research Scholar,

More information

Multilayer Perceptron

Multilayer Perceptron Outline Hong Chang Institute of Computing Technology, Chinese Academy of Sciences Machine Learning Methods (Fall 2012) Outline Outline I 1 Introduction 2 Single Perceptron 3 Boolean Function Learning 4

More information

Neural Networks (Part 1) Goals for the lecture

Neural Networks (Part 1) Goals for the lecture Neural Networks (Part ) Mark Craven and David Page Computer Sciences 760 Spring 208 www.biostat.wisc.edu/~craven/cs760/ Some of the slides in these lectures have been adapted/borrowed from materials developed

More information

From CDF to PDF A Density Estimation Method for High Dimensional Data

From CDF to PDF A Density Estimation Method for High Dimensional Data From CDF to PDF A Density Estimation Method for High Dimensional Data Shengdong Zhang Simon Fraser University sza75@sfu.ca arxiv:1804.05316v1 [stat.ml] 15 Apr 2018 April 17, 2018 1 Introduction Probability

More information

Neural networks. Chapter 19, Sections 1 5 1

Neural networks. Chapter 19, Sections 1 5 1 Neural networks Chapter 19, Sections 1 5 Chapter 19, Sections 1 5 1 Outline Brains Neural networks Perceptrons Multilayer perceptrons Applications of neural networks Chapter 19, Sections 1 5 2 Brains 10

More information

Nonlinear System Identification Based on a Novel Adaptive Fuzzy Wavelet Neural Network

Nonlinear System Identification Based on a Novel Adaptive Fuzzy Wavelet Neural Network Nonlinear System Identification Based on a Novel Adaptive Fuzzy Wavelet Neural Network Maryam Salimifard*, Ali Akbar Safavi** *School of Electrical Engineering, Amirkabir University of Technology, Tehran,

More information

NONLINEAR CLASSIFICATION AND REGRESSION. J. Elder CSE 4404/5327 Introduction to Machine Learning and Pattern Recognition

NONLINEAR CLASSIFICATION AND REGRESSION. J. Elder CSE 4404/5327 Introduction to Machine Learning and Pattern Recognition NONLINEAR CLASSIFICATION AND REGRESSION Nonlinear Classification and Regression: Outline 2 Multi-Layer Perceptrons The Back-Propagation Learning Algorithm Generalized Linear Models Radial Basis Function

More information

Artificial Neural Networks (ANN) Xiaogang Su, Ph.D. Department of Mathematical Science University of Texas at El Paso

Artificial Neural Networks (ANN) Xiaogang Su, Ph.D. Department of Mathematical Science University of Texas at El Paso Artificial Neural Networks (ANN) Xiaogang Su, Ph.D. Department of Mathematical Science University of Texas at El Paso xsu@utep.edu Fall, 2018 Outline Introduction A Brief History ANN Architecture Terminology

More information

Artificial Neural Networks. MGS Lecture 2

Artificial Neural Networks. MGS Lecture 2 Artificial Neural Networks MGS 2018 - Lecture 2 OVERVIEW Biological Neural Networks Cell Topology: Input, Output, and Hidden Layers Functional description Cost functions Training ANNs Back-Propagation

More information

A FUZZY NEURAL NETWORK MODEL FOR FORECASTING STOCK PRICE

A FUZZY NEURAL NETWORK MODEL FOR FORECASTING STOCK PRICE A FUZZY NEURAL NETWORK MODEL FOR FORECASTING STOCK PRICE Li Sheng Institute of intelligent information engineering Zheiang University Hangzhou, 3007, P. R. China ABSTRACT In this paper, a neural network-driven

More information

Reading Group on Deep Learning Session 1

Reading Group on Deep Learning Session 1 Reading Group on Deep Learning Session 1 Stephane Lathuiliere & Pablo Mesejo 2 June 2016 1/31 Contents Introduction to Artificial Neural Networks to understand, and to be able to efficiently use, the popular

More information

Speaker Representation and Verification Part II. by Vasileios Vasilakakis

Speaker Representation and Verification Part II. by Vasileios Vasilakakis Speaker Representation and Verification Part II by Vasileios Vasilakakis Outline -Approaches of Neural Networks in Speaker/Speech Recognition -Feed-Forward Neural Networks -Training with Back-propagation

More information

Neural networks. Chapter 20, Section 5 1

Neural networks. Chapter 20, Section 5 1 Neural networks Chapter 20, Section 5 Chapter 20, Section 5 Outline Brains Neural networks Perceptrons Multilayer perceptrons Applications of neural networks Chapter 20, Section 5 2 Brains 0 neurons of

More information

Neural networks. Chapter 20. Chapter 20 1

Neural networks. Chapter 20. Chapter 20 1 Neural networks Chapter 20 Chapter 20 1 Outline Brains Neural networks Perceptrons Multilayer networks Applications of neural networks Chapter 20 2 Brains 10 11 neurons of > 20 types, 10 14 synapses, 1ms

More information

Neural Networks. Chapter 18, Section 7. TB Artificial Intelligence. Slides from AIMA 1/ 21

Neural Networks. Chapter 18, Section 7. TB Artificial Intelligence. Slides from AIMA   1/ 21 Neural Networks Chapter 8, Section 7 TB Artificial Intelligence Slides from AIMA http://aima.cs.berkeley.edu / 2 Outline Brains Neural networks Perceptrons Multilayer perceptrons Applications of neural

More information

Deep Feedforward Networks

Deep Feedforward Networks Deep Feedforward Networks Liu Yang March 30, 2017 Liu Yang Short title March 30, 2017 1 / 24 Overview 1 Background A general introduction Example 2 Gradient based learning Cost functions Output Units 3

More information

APPLICATION OF RADIAL BASIS FUNCTION NEURAL NETWORK, TO ESTIMATE THE STATE OF HEALTH FOR LFP BATTERY

APPLICATION OF RADIAL BASIS FUNCTION NEURAL NETWORK, TO ESTIMATE THE STATE OF HEALTH FOR LFP BATTERY International Journal of Electrical and Electronics Engineering (IJEEE) ISSN(P): 2278-9944; ISSN(E): 2278-9952 Vol. 7, Issue 1, Dec - Jan 2018, 1-6 IASET APPLICATION OF RADIAL BASIS FUNCTION NEURAL NETWORK,

More information

CS:4420 Artificial Intelligence

CS:4420 Artificial Intelligence CS:4420 Artificial Intelligence Spring 2018 Neural Networks Cesare Tinelli The University of Iowa Copyright 2004 18, Cesare Tinelli and Stuart Russell a a These notes were originally developed by Stuart

More information

Artificial Neural Networks D B M G. Data Base and Data Mining Group of Politecnico di Torino. Elena Baralis. Politecnico di Torino

Artificial Neural Networks D B M G. Data Base and Data Mining Group of Politecnico di Torino. Elena Baralis. Politecnico di Torino Artificial Neural Networks Data Base and Data Mining Group of Politecnico di Torino Elena Baralis Politecnico di Torino Artificial Neural Networks Inspired to the structure of the human brain Neurons as

More information

CSE 417T: Introduction to Machine Learning. Final Review. Henry Chai 12/4/18

CSE 417T: Introduction to Machine Learning. Final Review. Henry Chai 12/4/18 CSE 417T: Introduction to Machine Learning Final Review Henry Chai 12/4/18 Overfitting Overfitting is fitting the training data more than is warranted Fitting noise rather than signal 2 Estimating! "#$

More information

1983. Gearbox fault diagnosis based on local mean decomposition, permutation entropy and extreme learning machine

1983. Gearbox fault diagnosis based on local mean decomposition, permutation entropy and extreme learning machine 1983. Gearbox fault diagnosis based on local mean decomposition, permutation entropy and extreme learning machine Yu Wei 1, Minqiang Xu 2, Yongbo Li 3, Wenhu Huang 4 Department of Astronautical Science

More information

Deep Learning Architecture for Univariate Time Series Forecasting

Deep Learning Architecture for Univariate Time Series Forecasting CS229,Technical Report, 2014 Deep Learning Architecture for Univariate Time Series Forecasting Dmitry Vengertsev 1 Abstract This paper studies the problem of applying machine learning with deep architecture

More information

Short-term wind forecasting using artificial neural networks (ANNs)

Short-term wind forecasting using artificial neural networks (ANNs) Energy and Sustainability II 197 Short-term wind forecasting using artificial neural networks (ANNs) M. G. De Giorgi, A. Ficarella & M. G. Russo Department of Engineering Innovation, Centro Ricerche Energia

More information

CSE 352 (AI) LECTURE NOTES Professor Anita Wasilewska. NEURAL NETWORKS Learning

CSE 352 (AI) LECTURE NOTES Professor Anita Wasilewska. NEURAL NETWORKS Learning CSE 352 (AI) LECTURE NOTES Professor Anita Wasilewska NEURAL NETWORKS Learning Neural Networks Classifier Short Presentation INPUT: classification data, i.e. it contains an classification (class) attribute.

More information

Need for Deep Networks Perceptron. Can only model linear functions. Kernel Machines. Non-linearity provided by kernels

Need for Deep Networks Perceptron. Can only model linear functions. Kernel Machines. Non-linearity provided by kernels Need for Deep Networks Perceptron Can only model linear functions Kernel Machines Non-linearity provided by kernels Need to design appropriate kernels (possibly selecting from a set, i.e. kernel learning)

More information

Support Vector Regression (SVR) Descriptions of SVR in this discussion follow that in Refs. (2, 6, 7, 8, 9). The literature

Support Vector Regression (SVR) Descriptions of SVR in this discussion follow that in Refs. (2, 6, 7, 8, 9). The literature Support Vector Regression (SVR) Descriptions of SVR in this discussion follow that in Refs. (2, 6, 7, 8, 9). The literature suggests the design variables should be normalized to a range of [-1,1] or [0,1].

More information

Least Square-Support Vector Machine Based Modeling for Electric Arc Furnaces

Least Square-Support Vector Machine Based Modeling for Electric Arc Furnaces Least Square-Support Vector Machine Based Modeling for Electric Arc Furnaces by Leyla Vakilian Bachelor of Computer Engineering, Shahid Beheshti University, 2001 A thesis Presented to Ryerson University

More information

The Application of Extreme Learning Machine based on Gaussian Kernel in Image Classification

The Application of Extreme Learning Machine based on Gaussian Kernel in Image Classification he Application of Extreme Learning Machine based on Gaussian Kernel in Image Classification Weijie LI, Yi LIN Postgraduate student in College of Survey and Geo-Informatics, tongji university Email: 1633289@tongji.edu.cn

More information

ECLT 5810 Classification Neural Networks. Reference: Data Mining: Concepts and Techniques By J. Hand, M. Kamber, and J. Pei, Morgan Kaufmann

ECLT 5810 Classification Neural Networks. Reference: Data Mining: Concepts and Techniques By J. Hand, M. Kamber, and J. Pei, Morgan Kaufmann ECLT 5810 Classification Neural Networks Reference: Data Mining: Concepts and Techniques By J. Hand, M. Kamber, and J. Pei, Morgan Kaufmann Neural Networks A neural network is a set of connected input/output

More information

Cheng Soon Ong & Christian Walder. Canberra February June 2018

Cheng Soon Ong & Christian Walder. Canberra February June 2018 Cheng Soon Ong & Christian Walder Research Group and College of Engineering and Computer Science Canberra February June 2018 Outlines Overview Introduction Linear Algebra Probability Linear Regression

More information

Address for Correspondence

Address for Correspondence Research Article APPLICATION OF ARTIFICIAL NEURAL NETWORK FOR INTERFERENCE STUDIES OF LOW-RISE BUILDINGS 1 Narayan K*, 2 Gairola A Address for Correspondence 1 Associate Professor, Department of Civil

More information

Keywords- Source coding, Huffman encoding, Artificial neural network, Multilayer perceptron, Backpropagation algorithm

Keywords- Source coding, Huffman encoding, Artificial neural network, Multilayer perceptron, Backpropagation algorithm Volume 4, Issue 5, May 2014 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com Huffman Encoding

More information

Last update: October 26, Neural networks. CMSC 421: Section Dana Nau

Last update: October 26, Neural networks. CMSC 421: Section Dana Nau Last update: October 26, 207 Neural networks CMSC 42: Section 8.7 Dana Nau Outline Applications of neural networks Brains Neural network units Perceptrons Multilayer perceptrons 2 Example Applications

More information

Neural Networks. Single-layer neural network. CSE 446: Machine Learning Emily Fox University of Washington March 10, /9/17

Neural Networks. Single-layer neural network. CSE 446: Machine Learning Emily Fox University of Washington March 10, /9/17 3/9/7 Neural Networks Emily Fox University of Washington March 0, 207 Slides adapted from Ali Farhadi (via Carlos Guestrin and Luke Zettlemoyer) Single-layer neural network 3/9/7 Perceptron as a neural

More information

Engineering Part IIB: Module 4F10 Statistical Pattern Processing Lecture 6: Multi-Layer Perceptrons I

Engineering Part IIB: Module 4F10 Statistical Pattern Processing Lecture 6: Multi-Layer Perceptrons I Engineering Part IIB: Module 4F10 Statistical Pattern Processing Lecture 6: Multi-Layer Perceptrons I Phil Woodland: pcw@eng.cam.ac.uk Michaelmas 2012 Engineering Part IIB: Module 4F10 Introduction In

More information

Knowledge Extraction from DBNs for Images

Knowledge Extraction from DBNs for Images Knowledge Extraction from DBNs for Images Son N. Tran and Artur d Avila Garcez Department of Computer Science City University London Contents 1 Introduction 2 Knowledge Extraction from DBNs 3 Experimental

More information

N. Sarikaya Department of Aircraft Electrical and Electronics Civil Aviation School Erciyes University Kayseri 38039, Turkey

N. Sarikaya Department of Aircraft Electrical and Electronics Civil Aviation School Erciyes University Kayseri 38039, Turkey Progress In Electromagnetics Research B, Vol. 6, 225 237, 2008 ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM FOR THE COMPUTATION OF THE CHARACTERISTIC IMPEDANCE AND THE EFFECTIVE PERMITTIVITY OF THE MICRO-COPLANAR

More information

Introduction to Neural Networks

Introduction to Neural Networks CUONG TUAN NGUYEN SEIJI HOTTA MASAKI NAKAGAWA Tokyo University of Agriculture and Technology Copyright by Nguyen, Hotta and Nakagawa 1 Pattern classification Which category of an input? Example: Character

More information

ARTIFICIAL NEURAL NETWORK PART I HANIEH BORHANAZAD

ARTIFICIAL NEURAL NETWORK PART I HANIEH BORHANAZAD ARTIFICIAL NEURAL NETWORK PART I HANIEH BORHANAZAD WHAT IS A NEURAL NETWORK? The simplest definition of a neural network, more properly referred to as an 'artificial' neural network (ANN), is provided

More information

Supervised (BPL) verses Hybrid (RBF) Learning. By: Shahed Shahir

Supervised (BPL) verses Hybrid (RBF) Learning. By: Shahed Shahir Supervised (BPL) verses Hybrid (RBF) Learning By: Shahed Shahir 1 Outline I. Introduction II. Supervised Learning III. Hybrid Learning IV. BPL Verses RBF V. Supervised verses Hybrid learning VI. Conclusion

More information

Neural Networks, Computation Graphs. CMSC 470 Marine Carpuat

Neural Networks, Computation Graphs. CMSC 470 Marine Carpuat Neural Networks, Computation Graphs CMSC 470 Marine Carpuat Binary Classification with a Multi-layer Perceptron φ A = 1 φ site = 1 φ located = 1 φ Maizuru = 1 φ, = 2 φ in = 1 φ Kyoto = 1 φ priest = 0 φ

More information

Confidence Estimation Methods for Neural Networks: A Practical Comparison

Confidence Estimation Methods for Neural Networks: A Practical Comparison , 6-8 000, Confidence Estimation Methods for : A Practical Comparison G. Papadopoulos, P.J. Edwards, A.F. Murray Department of Electronics and Electrical Engineering, University of Edinburgh Abstract.

More information

Machine Learning. Neural Networks. (slides from Domingos, Pardo, others)

Machine Learning. Neural Networks. (slides from Domingos, Pardo, others) Machine Learning Neural Networks (slides from Domingos, Pardo, others) Human Brain Neurons Input-Output Transformation Input Spikes Output Spike Spike (= a brief pulse) (Excitatory Post-Synaptic Potential)

More information

A recursive algorithm based on the extended Kalman filter for the training of feedforward neural models. Isabelle Rivals and Léon Personnaz

A recursive algorithm based on the extended Kalman filter for the training of feedforward neural models. Isabelle Rivals and Léon Personnaz In Neurocomputing 2(-3): 279-294 (998). A recursive algorithm based on the extended Kalman filter for the training of feedforward neural models Isabelle Rivals and Léon Personnaz Laboratoire d'électronique,

More information

Short Term Load Forecasting Based Artificial Neural Network

Short Term Load Forecasting Based Artificial Neural Network Short Term Load Forecasting Based Artificial Neural Network Dr. Adel M. Dakhil Department of Electrical Engineering Misan University Iraq- Misan Dr.adelmanaa@gmail.com Abstract Present study develops short

More information

Neural Networks and the Back-propagation Algorithm

Neural Networks and the Back-propagation Algorithm Neural Networks and the Back-propagation Algorithm Francisco S. Melo In these notes, we provide a brief overview of the main concepts concerning neural networks and the back-propagation algorithm. We closely

More information

Journal of Engineering Science and Technology Review 7 (1) (2014)

Journal of Engineering Science and Technology Review 7 (1) (2014) Jestr Journal of Engineering Science and Technology Review 7 () (204) 32 36 JOURNAL OF Engineering Science and Technology Review www.jestr.org Particle Swarm Optimization-based BP Neural Network for UHV

More information

Nonlinear Classification

Nonlinear Classification Nonlinear Classification INFO-4604, Applied Machine Learning University of Colorado Boulder October 5-10, 2017 Prof. Michael Paul Linear Classification Most classifiers we ve seen use linear functions

More information

Machine Learning. Neural Networks. (slides from Domingos, Pardo, others)

Machine Learning. Neural Networks. (slides from Domingos, Pardo, others) Machine Learning Neural Networks (slides from Domingos, Pardo, others) For this week, Reading Chapter 4: Neural Networks (Mitchell, 1997) See Canvas For subsequent weeks: Scaling Learning Algorithms toward

More information

MODELLING OF TOOL LIFE, TORQUE AND THRUST FORCE IN DRILLING: A NEURO-FUZZY APPROACH

MODELLING OF TOOL LIFE, TORQUE AND THRUST FORCE IN DRILLING: A NEURO-FUZZY APPROACH ISSN 1726-4529 Int j simul model 9 (2010) 2, 74-85 Original scientific paper MODELLING OF TOOL LIFE, TORQUE AND THRUST FORCE IN DRILLING: A NEURO-FUZZY APPROACH Roy, S. S. Department of Mechanical Engineering,

More information

Hand Written Digit Recognition using Kalman Filter

Hand Written Digit Recognition using Kalman Filter International Journal of Electronics and Communication Engineering. ISSN 0974-2166 Volume 5, Number 4 (2012), pp. 425-434 International Research Publication House http://www.irphouse.com Hand Written Digit

More information

Artificial Neural Networks" and Nonparametric Methods" CMPSCI 383 Nov 17, 2011!

Artificial Neural Networks and Nonparametric Methods CMPSCI 383 Nov 17, 2011! Artificial Neural Networks" and Nonparametric Methods" CMPSCI 383 Nov 17, 2011! 1 Todayʼs lecture" How the brain works (!)! Artificial neural networks! Perceptrons! Multilayer feed-forward networks! Error

More information

COMP 551 Applied Machine Learning Lecture 14: Neural Networks

COMP 551 Applied Machine Learning Lecture 14: Neural Networks COMP 551 Applied Machine Learning Lecture 14: Neural Networks Instructor: Ryan Lowe (ryan.lowe@mail.mcgill.ca) Slides mostly by: Class web page: www.cs.mcgill.ca/~hvanho2/comp551 Unless otherwise noted,

More information

Data Quality in Hybrid Neuro-Fuzzy based Soft-Sensor Models: An Experimental Study

Data Quality in Hybrid Neuro-Fuzzy based Soft-Sensor Models: An Experimental Study IAENG International Journal of Computer Science, 37:1, IJCS_37_1_8 Data Quality in Hybrid Neuro-Fuzzy based Soft-Sensor Models: An Experimental Study S. Jassar, Student Member, IEEE, Z. Liao, Member, ASHRAE,

More information

Choosing Variables with a Genetic Algorithm for Econometric models based on Neural Networks learning and adaptation.

Choosing Variables with a Genetic Algorithm for Econometric models based on Neural Networks learning and adaptation. Choosing Variables with a Genetic Algorithm for Econometric models based on Neural Networks learning and adaptation. Daniel Ramírez A., Israel Truijillo E. LINDA LAB, Computer Department, UNAM Facultad

More information

Supporting Information

Supporting Information Supporting Information Convolutional Embedding of Attributed Molecular Graphs for Physical Property Prediction Connor W. Coley a, Regina Barzilay b, William H. Green a, Tommi S. Jaakkola b, Klavs F. Jensen

More information

Analysis of Multilayer Neural Network Modeling and Long Short-Term Memory

Analysis of Multilayer Neural Network Modeling and Long Short-Term Memory Analysis of Multilayer Neural Network Modeling and Long Short-Term Memory Danilo López, Nelson Vera, Luis Pedraza International Science Index, Mathematical and Computational Sciences waset.org/publication/10006216

More information

Neural Network Training

Neural Network Training Neural Network Training Sargur Srihari Topics in Network Training 0. Neural network parameters Probabilistic problem formulation Specifying the activation and error functions for Regression Binary classification

More information

C4 Phenomenological Modeling - Regression & Neural Networks : Computational Modeling and Simulation Instructor: Linwei Wang

C4 Phenomenological Modeling - Regression & Neural Networks : Computational Modeling and Simulation Instructor: Linwei Wang C4 Phenomenological Modeling - Regression & Neural Networks 4040-849-03: Computational Modeling and Simulation Instructor: Linwei Wang Recall.. The simple, multiple linear regression function ŷ(x) = a

More information

Pattern Recognition and Machine Learning

Pattern Recognition and Machine Learning Christopher M. Bishop Pattern Recognition and Machine Learning ÖSpri inger Contents Preface Mathematical notation Contents vii xi xiii 1 Introduction 1 1.1 Example: Polynomial Curve Fitting 4 1.2 Probability

More information

STA 4273H: Statistical Machine Learning

STA 4273H: Statistical Machine Learning STA 4273H: Statistical Machine Learning Russ Salakhutdinov Department of Statistics! rsalakhu@utstat.toronto.edu! http://www.utstat.utoronto.ca/~rsalakhu/ Sidney Smith Hall, Room 6002 Lecture 3 Linear

More information

<Special Topics in VLSI> Learning for Deep Neural Networks (Back-propagation)

<Special Topics in VLSI> Learning for Deep Neural Networks (Back-propagation) Learning for Deep Neural Networks (Back-propagation) Outline Summary of Previous Standford Lecture Universal Approximation Theorem Inference vs Training Gradient Descent Back-Propagation

More information

Acceleration of Levenberg-Marquardt method training of chaotic systems fuzzy modeling

Acceleration of Levenberg-Marquardt method training of chaotic systems fuzzy modeling ISSN 746-7233, England, UK World Journal of Modelling and Simulation Vol. 3 (2007) No. 4, pp. 289-298 Acceleration of Levenberg-Marquardt method training of chaotic systems fuzzy modeling Yuhui Wang, Qingxian

More information

Artificial Neural Network

Artificial Neural Network Artificial Neural Network Contents 2 What is ANN? Biological Neuron Structure of Neuron Types of Neuron Models of Neuron Analogy with human NN Perceptron OCR Multilayer Neural Network Back propagation

More information

Lecture 6. Regression

Lecture 6. Regression Lecture 6. Regression Prof. Alan Yuille Summer 2014 Outline 1. Introduction to Regression 2. Binary Regression 3. Linear Regression; Polynomial Regression 4. Non-linear Regression; Multilayer Perceptron

More information

Research Article Design Feed Forward Neural Network to Solve Singular Boundary Value Problems

Research Article Design Feed Forward Neural Network to Solve Singular Boundary Value Problems ISRN Applied Mathematics Volume 2013, Article ID 650467, 7 pages http://d.doi.org/10.1155/2013/650467 Research Article Design Feed Forward Neural Network to Solve Singular Boundary Value Problems Luma

More information

A STATE-SPACE NEURAL NETWORK FOR MODELING DYNAMICAL NONLINEAR SYSTEMS

A STATE-SPACE NEURAL NETWORK FOR MODELING DYNAMICAL NONLINEAR SYSTEMS A STATE-SPACE NEURAL NETWORK FOR MODELING DYNAMICAL NONLINEAR SYSTEMS Karima Amoura Patrice Wira and Said Djennoune Laboratoire CCSP Université Mouloud Mammeri Tizi Ouzou Algeria Laboratoire MIPS Université

More information

18.6 Regression and Classification with Linear Models

18.6 Regression and Classification with Linear Models 18.6 Regression and Classification with Linear Models 352 The hypothesis space of linear functions of continuous-valued inputs has been used for hundreds of years A univariate linear function (a straight

More information

Day 3 Lecture 3. Optimizing deep networks

Day 3 Lecture 3. Optimizing deep networks Day 3 Lecture 3 Optimizing deep networks Convex optimization A function is convex if for all α [0,1]: f(x) Tangent line Examples Quadratics 2-norms Properties Local minimum is global minimum x Gradient

More information

Artificial Neural Networks. Edward Gatt

Artificial Neural Networks. Edward Gatt Artificial Neural Networks Edward Gatt What are Neural Networks? Models of the brain and nervous system Highly parallel Process information much more like the brain than a serial computer Learning Very

More information

EM-algorithm for Training of State-space Models with Application to Time Series Prediction

EM-algorithm for Training of State-space Models with Application to Time Series Prediction EM-algorithm for Training of State-space Models with Application to Time Series Prediction Elia Liitiäinen, Nima Reyhani and Amaury Lendasse Helsinki University of Technology - Neural Networks Research

More information

The Perceptron. Volker Tresp Summer 2016

The Perceptron. Volker Tresp Summer 2016 The Perceptron Volker Tresp Summer 2016 1 Elements in Learning Tasks Collection, cleaning and preprocessing of training data Definition of a class of learning models. Often defined by the free model parameters

More information

Using a Hopfield Network: A Nuts and Bolts Approach

Using a Hopfield Network: A Nuts and Bolts Approach Using a Hopfield Network: A Nuts and Bolts Approach November 4, 2013 Gershon Wolfe, Ph.D. Hopfield Model as Applied to Classification Hopfield network Training the network Updating nodes Sequencing of

More information

Nonlinear system modeling with deep neural networks and autoencoders algorithm

Nonlinear system modeling with deep neural networks and autoencoders algorithm Nonlinear system modeling with deep neural networks and autoencoders algorithm Erick De la Rosa, Wen Yu Departamento de Control Automatico CINVESTAV-IPN Mexico City, Mexico yuw@ctrl.cinvestav.mx Xiaoou

More information

A New Weight Initialization using Statistically Resilient Method and Moore-Penrose Inverse Method for SFANN

A New Weight Initialization using Statistically Resilient Method and Moore-Penrose Inverse Method for SFANN A New Weight Initialization using Statistically Resilient Method and Moore-Penrose Inverse Method for SFANN Apeksha Mittal, Amit Prakash Singh and Pravin Chandra University School of Information and Communication

More information

Application of Artificial Neural Networks in Evaluation and Identification of Electrical Loss in Transformers According to the Energy Consumption

Application of Artificial Neural Networks in Evaluation and Identification of Electrical Loss in Transformers According to the Energy Consumption Application of Artificial Neural Networks in Evaluation and Identification of Electrical Loss in Transformers According to the Energy Consumption ANDRÉ NUNES DE SOUZA, JOSÉ ALFREDO C. ULSON, IVAN NUNES

More information

Learning from Examples

Learning from Examples Learning from Examples Data fitting Decision trees Cross validation Computational learning theory Linear classifiers Neural networks Nonparametric methods: nearest neighbor Support vector machines Ensemble

More information

Machine Learning and Data Mining. Multi-layer Perceptrons & Neural Networks: Basics. Prof. Alexander Ihler

Machine Learning and Data Mining. Multi-layer Perceptrons & Neural Networks: Basics. Prof. Alexander Ihler + Machine Learning and Data Mining Multi-layer Perceptrons & Neural Networks: Basics Prof. Alexander Ihler Linear Classifiers (Perceptrons) Linear Classifiers a linear classifier is a mapping which partitions

More information

International Journal of Advanced Research in Computer Science and Software Engineering

International Journal of Advanced Research in Computer Science and Software Engineering Volume 3, Issue 4, April 2013 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com Application of

More information

A new method for short-term load forecasting based on chaotic time series and neural network

A new method for short-term load forecasting based on chaotic time series and neural network A new method for short-term load forecasting based on chaotic time series and neural network Sajjad Kouhi*, Navid Taghizadegan Electrical Engineering Department, Azarbaijan Shahid Madani University, Tabriz,

More information

Optimization Approximation Solution for Regression Problem Based on Extremal Learning Machine

Optimization Approximation Solution for Regression Problem Based on Extremal Learning Machine Optimization Approximation Solution for Regression Problem Based on Extremal Learning Machine Yubo Yuan Yuguang Wang Feilong Cao Department of Mathematics, China Jiliang University, Hangzhou 3008, Zhejiang

More information