Search results for "Backpropagation"
showing 10 items of 27 documents
Detection and classification of microcalcifications clusters in digitized mammograms
2005
In the present paper we discuss a new approach for the detection of microcalcification clusters, based on neural networks and developed as part of the MAGIC-5 project, an INFN-funded program which aims at the development and implementation of CAD algorithms in a GRID-based distributed environment. The proposed approach has as its roots the desire to maximize the rejection of background during the analytical pre-processing stage, in order to train and test the neural network with as clean as possible a sample and therefore maximize its performance. The algorithm is composed of three modules: the image pre-processing, the feature extraction component and the Backpropagation Neural Network mod…
A system based on neural architectures for the reconstruction of 3-D shapes from images
1991
The connectionist approach to the recovery of 3-D shape information from 2-D images developed by the authors, is based on a system made up by two cascaded neural networks. The first network is an implementation of the BCS, an architecture which derives from a biological model of the low level visual processes developed by Grossberg and Mingolla: this architecture extracts a sort of brightness gradient map from the image. The second network is a backpropagation architecture that supplies an estimate of the geometric parameters of the objects in the scene under consideration, starting from the outputs of the BCS. A detailed description of the system and the experimental results obtained by si…
Deep CNN-ELM Hybrid Models for Fire Detection in Images
2018
In this paper, we propose a hybrid model consisting of a Deep Convolutional feature extractor followed by a fast and accurate classifier, the Extreme Learning Machine, for the purpose of fire detection in images. The reason behind using such a model is that Deep CNNs used for image classification take a very long time to train. Even with pre-trained models, the fully connected layers need to be trained with backpropagation, which can be very slow. In contrast, we propose to employ the Extreme Learning Machine (ELM) as the final classifier trained on pre-trained Deep CNN feature extractor. We apply this hybrid model on the problem of fire detection in images. We use state of the art Deep CNN…
Multi-step ahead electricity price forecasting using a hybrid model based on two-layer decomposition technique and BP neural network optimized by fir…
2017
In the deregulated competitive electricity market, the price which reflects the relationship between electricity supply and demand is one of the most important elements, making it crucial for all market participants to precisely forecast the electricity price. However, electricity price series usually has complex features such as non-linearity, non-stationarity and volatility, which makes the price forecasting turn out to be very difficult. In order to improve the accuracy of electricity price forecasting, this paper first proposes a two-layer decomposition technique and then develops a hybrid model based on fast ensemble empirical mode decomposition (FEEMD), variational mode decomposition …
A machine learning approach to determine airport asphalt concrete layer moduli using heavy weight deflectometer data
2021
An integrated approach based on machine learning and data augmentation techniques has been developed in order to predict the stiffness modulus of the asphalt concrete layer of an airport runway, from data acquired with a heavy weight deflectometer (HWD). The predictive model relies on a shallow neural network (SNN) trained with the results of a backcalculation, by means of a data augmentation method and can produce estimations of the stiffness modulus even at runway points not yet sampled. The Bayesian regularization algorithm was used for training of the feedforward backpropagation SNN, and a k-fold cross-validation procedure was implemented for a fair performance evaluation. The testing p…
Architectural improvements and FPGA implementation of a multimodel neuroprocessor
2003
Since neural networks (NNs) require an enormous amount of learning time, various kinds of dedicated parallel computers have been developed. In the paper a 2-D systolic array (SA) of dedicated processing elements (PEs) also called systolic cells (SCs) is presented as the heart of a multimodel neural-network accelerator. The instruction set of the SA allows the implementation of several neural algorithms, including error back propagation and a self organizing feature map algorithm. Several special architectural facilities are presented in the paper in order to improve the 2-D SA performance. A swapping mechanism of the weight matrix allows the implementation of NNs larger than 2-D SA. A systo…
Path relinking and GRG for artificial neural networks
2006
Artificial neural networks (ANN) have been widely used for both classification and prediction. This paper is focused on the prediction problem in which an unknown function is approximated. ANNs can be viewed as models of real systems, built by tuning parameters known as weights. In training the net, the problem is to find the weights that optimize its performance (i.e., to minimize the error over the training set). Although the most popular method for training these networks is back propagation, other optimization methods such as tabu search or scatter search have been successfully applied to solve this problem. In this paper we propose a path relinking implementation to solve the neural ne…
The Prediction of Human Intestinal Absorption Based on the Molecular Structure
2014
Human Intestinal Absorption (HIA) has been modeled many times by using classification models. However, regression models are scarce. Here, Artificial Neural Networks (ANNs) are implemented for this purpose. A dataset of structurally diverse chemicals with their respective experimental HIA were used to design robust, true predictive and widespread applicable ANN models. An input variables pool was made up of structural invariants calculated by using either Dragon or our software Desmol 1. The selection of best variables was performed following three steps using the entire dataset of molecules. Firstly, variables poorly correlated with the experimental data were eliminated. Secondly, input va…
A Neural Architecture for Segmentation and Modelling of Range Data
2003
A novel, two stage, neural architecture for the segmentation of range data and their modeling with undeformed superquadrics is presented. The system is composed by two distinct neural stages: a SOM is used to perform data segmentation, and, for each segment, a multi-layer feed-forward network performs model estimation. The topology preserving nature of the SOM algorithm makes this architecture suited to cluster data with respect to sudden curvature variations. The second stage is designed to model and compute the inside-outside function of an undeformed superquadric in whatever attitude, starting form the (x, y, z) data triples. The network has been trained using backpropagation, and the we…
Forecasting daily urban electric load profiles using artificial neural networks
2004
The paper illustrates a combined approach based on unsupervised and supervised neural networks for the electric energy demand forecasting of a suburban area with a prediction time of 24 h. A preventive classification of the historical load data is performed during the unsupervised stage by means of a Kohonen's self organizing map (SOM). The actual forecast is obtained using a two layered feed forward neural network, trained with the back propagation with momentum learning algorithm. In order to investigate the influence of climate variability on the electricity consumption, the neural network is trained using weather data (temperature, relative humidity, global solar radiation) along with h…