Feedforward neural networks with random weights

Schmidt, Walter; Kraaijveld, Martin; Duin, Robert P. W.

doi:10.1109/icpr.1992.201708

Cited by 336 publications

(211 citation statements)

References 3 publications

Supporting

Mentioning

201

Contrasting

Unclassified

Order By: Relevance

“…In the first step we collect social and academic data from a set of computer science students. Using such dataset, we designed a classifier with reject option where a Feedforward Neural Network with Random Weights (FNNRW) [Schmidt et al 1992] was used as a base learner.…”

Section: Methodsmentioning

confidence: 99%

A machine learning approach to identify and prioritize college students at risk of dropping out

Barbosa

Santos

Pordeus

2017

Anais Do XXVIII Simpósio Brasileiro De Informática Na Educação (SBIE 2017)

View full text Add to dashboard Cite

Abstract. In this paper, we present a student dropout prediction strategy based on the classification with reject option paradigm. In such strategy, our method classifies students into dropout prone or non-dropout prone classes and may also reject classifying students when the algorithm does not provide a reliable prediction. The rejected students are the ones that could be classified into either class, and so are probably the ones with more chances of success when subjected to personalized intervention activities. In the proposed method, the reject zone can be adjusted so that the number of rejected students can meet the available workforce of the educational institution. Our method was tested on a dataset collected from 892 undergraduate students from 2005 to 2016.

show abstract

Section: Methodsmentioning

confidence: 99%

A machine learning approach to identify and prioritize college students at risk of dropping out

Barbosa

Santos

Pordeus

2017

Anais Do XXVIII Simpósio Brasileiro De Informática Na Educação (SBIE 2017)

View full text Add to dashboard Cite

show abstract

“…Predecessors of these models were proposed in a number of early works on feedforward architectures, e.g. in [42,43]. A more mature version of RW-FNNs, called Random Vector Functional-Link (RVFL) networks were introduced in [44,45].…”

Section: Related Workmentioning

confidence: 99%

Training neural networks on high-dimensional data using random projection

Wójcik

Kurdziel

2018

Pattern Anal Applic

View full text Add to dashboard Cite

Training deep neural networks (DNNs) on high-dimensional data with no spatial structure poses a major computational problem. It implies a network architecture with a huge input layer, which greatly increases the number of weights, often making the training infeasible. One solution to this problem is to reduce the dimensionality of the input space to a manageable size, and then train a deep network on a representation with fewer dimensions. Here, we focus on performing the dimensionality reduction step by randomly projecting the input data into a lower-dimensional space. Conceptually, this is equivalent to adding a random projection (RP) layer in front of the network. We study two variants of RP layers: one where the weights are fixed, and one where they are fine-tuned during network training. We evaluate the performance of DNNs with input layers constructed using several recently proposed RP schemes. These include: Gaussian, Achlioptas', Li's, subsampled randomized Hadamard transform (SRHT) and Count Sketch-based constructions. Our results demonstrate that DNNs with RP layer achieve competitive performance on high-dimensional real-world datasets. In particular, we show that SRHT and Count Sketch-based projections provide the best balance between the projection time and the network performance.

show abstract

“…Only the outputs of the unit are connected to the units of next layer. Therefore there is no feedback in the system [7]. …”

Section: Feed Forward Neural Networkmentioning

confidence: 99%

Using Neural Networks to Predict Secondary Structure for Protein Folding

Ibrahim¹,

Yasseen²

2017

JCC

View full text Add to dashboard Cite

Protein Secondary Structure Prediction (PSSP) is considered as one of the major challenging tasks in bioinformatics, so many solutions have been proposed to solve that problem via trying to achieve more accurate prediction results. The goal of this paper is to develop and implement an intelligent based system to predict secondary structure of a protein from its primary amino acid sequence by using five models of Neural Network (NN). These models are Feed Forward Neural Network (FNN), Learning Vector Quantization (LVQ), Probabilistic Neural Network (PNN), Convolutional Neural Network (CNN), and CNN Fine Tuning for PSSP. To evaluate our approaches two datasets have been used. The first one contains 114 protein samples, and the second one contains 1845 protein samples.

show abstract

Feedforward neural networks with random weights

Cited by 336 publications

References 3 publications

A machine learning approach to identify and prioritize college students at risk of dropping out

A machine learning approach to identify and prioritize college students at risk of dropping out

Training neural networks on high-dimensional data using random projection

Using Neural Networks to Predict Secondary Structure for Protein Folding

Contact Info

Product

Resources

About