scholarly journals Bias in Deep Neural Networks in Land Use Characterization for International Development

2021 ◽  
Vol 13 (15) ◽  
pp. 2908
Author(s):  
Do-Hyung Kim ◽  
Guzmán López ◽  
Diego Kiedanski ◽  
Iyke Maduako ◽  
Braulio Ríos ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Land Use ◽  
International Development ◽  
Deep Neural Networks ◽  
Satellite Image ◽  
Fine Tuning ◽  
Training Dataset ◽  
Model Accuracy ◽  
The Neural Network

Understanding the biases in Deep Neural Networks (DNN) based algorithms is gaining paramount importance due to its increased applications on many real-world problems. A known problem of DNN penalizing the underrepresented population could undermine the efficacy of development projects dependent on data produced using DNN-based models. In spite of this, the problems of biases in DNN for Land Use and Land Cover Classification (LULCC) have not been a subject of many studies. In this study, we explore ways to quantify biases in DNN for land use with an example of identifying school buildings in Colombia from satellite imagery. We implement a DNN-based model by fine-tuning an existing, pre-trained model for school building identification. The model achieved overall 84% accuracy. Then, we used socioeconomic covariates to analyze possible biases in the learned representation. The retrained deep neural network was used to extract visual features (embeddings) from satellite image tiles. The embeddings were clustered into four subtypes of schools, and the accuracy of the neural network model was assessed for each cluster. The distributions of various socioeconomic covariates by clusters were analyzed to identify the links between the model accuracy and the aforementioned covariates. Our results indicate that the model accuracy is lowest (57%) where the characteristics of the landscape are predominantly related to poverty and remoteness, which confirms our original assumption on the heterogeneous performances of Artificial Intelligence (AI) algorithms and their biases. Based on our findings, we identify possible sources of bias and present suggestions on how to prepare a balanced training dataset that would result in less biased AI algorithms. The framework used in our study to better understand biases in DNN models would be useful when Machine Learning (ML) techniques are adopted in lieu of ground-based data collection for international development programs. Because such programs aim to solve issues of social inequality, MLs are only applicable when they are transparent and accountable.

scholarly journals Evaluation of Power Insulator Detection Efficiency with the Use of Limited Training Dataset

2020 ◽  
Vol 10 (6) ◽  
pp. 2104
Author(s):  
Michał Tomaszewski ◽  
Paweł Michalski ◽  
Jakub Osuchowski
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Object Detection ◽  
Convolutional Neural Network ◽  
Deep Neural Networks ◽  
Detection Efficiency ◽  
Training Data ◽  
Training Dataset ◽  
Training Set ◽  
Convolutional Network

This article presents an analysis of the effectiveness of object detection in digital images with the application of a limited quantity of input. The possibility of using a limited set of learning data was achieved by developing a detailed scenario of the task, which strictly defined the conditions of detector operation in the considered case of a convolutional neural network. The described solution utilizes known architectures of deep neural networks in the process of learning and object detection. The article presents comparisons of results from detecting the most popular deep neural networks while maintaining a limited training set composed of a specific number of selected images from diagnostic video. The analyzed input material was recorded during an inspection flight conducted along high-voltage lines. The object detector was built for a power insulator. The main contribution of the presented papier is the evidence that a limited training set (in our case, just 60 training frames) could be used for object detection, assuming an outdoor scenario with low variability of environmental conditions. The decision of which network will generate the best result for such a limited training set is not a trivial task. Conducted research suggests that the deep neural networks will achieve different levels of effectiveness depending on the amount of training data. The most beneficial results were obtained for two convolutional neural networks: the faster region-convolutional neural network (faster R-CNN) and the region-based fully convolutional network (R-FCN). Faster R-CNN reached the highest AP (average precision) at a level of 0.8 for 60 frames. The R-FCN model gained a worse AP result; however, it can be noted that the relationship between the number of input samples and the obtained results has a significantly lower influence than in the case of other CNN models, which, in the authors’ assessment, is a desired feature in the case of a limited training set.

OPTIMIZATION PROCESS ANALYSIS FOR HYPERPARAMETERS OF NEURAL NETWORK DATA PROCESSING STRUCTURES

2020 ◽  
pp. 3-10
Author(s):  
V. N. Gridin ◽  
I. A. Evdokimov ◽  
B. R. Salem ◽  
V. I. Solodovnikov
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Neural Networks ◽  
Process Analysis ◽  
Validation Process ◽  
The Neural Network ◽  
Source Data ◽  
The Neural Networks ◽  
Qualitative Characteristics ◽  
Setting Parameters

The analysis of key stages, implementation features and functioning principles of the neural networks, including deep neural networks, has been carried out. The problems of choosing the number of hidden elements, methods for the internal topology selection and setting parameters are considered. It is shown that in the training and validation process it is possible to control the capacity of a neural network and evaluate the qualitative characteristics of the constructed model. The issues of construction processes automation and hyperparameters optimization of the neural network structures are considered depending on the user's tasks and the available source data. A number of approaches based on the use of probabilistic programming, evolutionary algorithms, and recurrent neural networks are presented.

scholarly journals A HYBRID MODEL USING THE PRETRAINED BERT AND DEEP NEURAL NETWORKS WITH RICH FEATURE FOR EXTRACTIVE TEXT SUMMARIZATION

2021 ◽  
Vol 37 (2) ◽  
pp. 123-143
Author(s):  
Tuan Minh Luu ◽  
Huong Thanh Le ◽  
Tan Minh Hoang
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Neural Networks ◽  
Text Summarization ◽  
Training Dataset ◽  
Extractive Summarization ◽  
Input Text ◽  
Summarization System ◽  
Fully Connected

Deep neural networks have been applied successfully to extractive text summarization tasks with the accompany of large training datasets. However, when the training dataset is not large enough, these models reveal certain limitations that affect the quality of the system’s summary. In this paper, we propose an extractive summarization system basing on a Convolutional Neural Network and a Fully Connected network for sentence selection. The pretrained BERT multilingual model is used to generate embeddings vectors from the input text. These vectors are combined with TF-IDF values to produce the input of the text summarization system. Redundant sentences from the output summary are eliminated by the Maximal Marginal Relevance method. Our system is evaluated with both English and Vietnamese languages using CNN and Baomoi datasets, respectively. Experimental results show that our system achieves better results comparing to existing works using the same dataset. It confirms that our approach can be effectively applied to summarize both English and Vietnamese languages.

Endek Classification Based On GLCM Using Artificial Neural Networks with Adam Optimization

2020 ◽  
Vol 9 (2) ◽  
pp. 285
Author(s):  
Putu Wahyu Tirta Guna ◽  
Luh Arida Ayu Ayu Rahning Putri
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Networks ◽  
Back Propagation ◽  
Classification Problem ◽  
Model Accuracy ◽  
Computing Power ◽  
The Neural Network ◽  
Extraction Algorithm ◽  
Better Than

Not many people know that endek cloth itself has 4 known variances. .Nowadays. Computing and classification algorithm can be implemented to solve classification problem with respect to the features data as input. We can use this computing power to digitalize these endek pattern. The features extraction algorithm used in this research is GLCM. Where these data will act as input for the neural network model later. There is a lot of optimizer algorithm to use in back propagation phase. In this research we  prefer to use adam which is one of the newest and most popular optimizer algorithm. To compare its performace we also use SGD which is older and popular optimizer algorithm. Later we find that adam algorithm generate 33% accuracy which is better than what SGD algorithm give, it is 23% accuracy. Longer epoch also give affect for overall model accuracy.

PIPxResNet: Penalty Induced Prototype-Based eXplainable Residual Neural Network for Heartbeat Classification

2021 ◽  
Author(s):  
Deepankar Nankani ◽  
Rashmi Dutta Baruah
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Neural Networks ◽  
Early Stage ◽  
Black Box ◽  
Training Dataset ◽  
Medical Attention ◽  
Generative Adversarial Network ◽  
Heartbeat Classification ◽  
General Physicians

Abstract Early stage heartbeat classification using the electrocardiogram signals can prevent cardiovascular diseases that causes millions of deaths annually around the world. In the past, researchers have used deep neural networks to achieve significant performance for heartbeat classification but their black-box nature and prediction rationale limits real-world deployment. We propose a Penalty Induced Prototype based eXplainable Residual Neural Network (PIPxResNet) that addresses the black-box nature of deep neural networks. PIPxResNet encodes the temporal variations of heartbeats by employing pretrained residual neural network following the concept of task transfer learning. The algorithm further extracts prototypes that are most representative of the training dataset that explain model predictions to general physicians, making them clinically relevant. The prototypes of a particular class having close resemblance to other class prototypes are penalised and their contribution towards corresponding class is reduced. In addition, the classification performance is improved by synthesising regular and irregular heartbeats using a deep convolution conditional generative adversarial network. The proposed method can easily be adopted to other domains that requires explanations for the classification tasks. The PIPxResNet performs at par with existing state-of-the-art algorithms without compromising individual class performance when tested on four publicly available annotated datasets. The proposed model is capable to perform automated screening and provide medical attention by simulating a clinical decision support system for general physicians.

scholarly journals Exploring deep neural networks via layer-peeled model: Minority collapse in imbalanced training

2021 ◽  
Vol 118 (43) ◽  
pp. e2103091118
Author(s):  
Cong Fang ◽  
Hangfeng He ◽  
Qi Long ◽  
Weijie J. Su
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Learning ◽  
Deep Neural Networks ◽  
Model Minority ◽  
Tight Frame ◽  
Learning Models ◽  
The Neural Network ◽  
Long Time ◽  
Topmost Layer

In this paper, we introduce the Layer-Peeled Model, a nonconvex, yet analytically tractable, optimization program, in a quest to better understand deep neural networks that are trained for a sufficiently long time. As the name suggests, this model is derived by isolating the topmost layer from the remainder of the neural network, followed by imposing certain constraints separately on the two parts of the network. We demonstrate that the Layer-Peeled Model, albeit simple, inherits many characteristics of well-trained neural networks, thereby offering an effective tool for explaining and predicting common empirical patterns of deep-learning training. First, when working on class-balanced datasets, we prove that any solution to this model forms a simplex equiangular tight frame, which, in part, explains the recently discovered phenomenon of neural collapse [V. Papyan, X. Y. Han, D. L. Donoho, Proc. Natl. Acad. Sci. U.S.A. 117, 24652–24663 (2020)]. More importantly, when moving to the imbalanced case, our analysis of the Layer-Peeled Model reveals a hitherto-unknown phenomenon that we term Minority Collapse, which fundamentally limits the performance of deep-learning models on the minority classes. In addition, we use the Layer-Peeled Model to gain insights into how to mitigate Minority Collapse. Interestingly, this phenomenon is first predicted by the Layer-Peeled Model before being confirmed by our computational experiments.

scholarly journals Construction of a neural network energy function for protein physics

2021 ◽  
Author(s):  
Huan Yang ◽  
Zhaoping Xiong ◽  
Francesco Zonta
Keyword(s):  
Neural Network ◽  
Experimental Data ◽  
Neural Networks ◽  
Energy Function ◽  
Protein Sequence ◽  
Deep Neural Networks ◽  
Protein Structures ◽  
Learning Method ◽  
Local Structures ◽  
The Neural Network

AbstractClassical potentials are widely used to describe protein physics, due to their simplicity and accuracy, but they are continuously challenged as real applications become more demanding with time. Deep neural networks could help generating alternative ways of describing protein physics. Here we propose an unsupervised learning method to derive a neural network energy function for proteins. The energy function is a probability density model learned from plenty of 3D local structures which have been extensively explored by evolution. We tested this model on a few applications (assessment of protein structures, protein dynamics and protein sequence design), showing that the neural network can correctly recognize patterns in protein structures. In other words, the neural network learned some aspects of protein physics from experimental data.

scholarly journals Tri-net for Semi-Supervised Deep Learning

2018 ◽  
Author(s):  
Dong-Dong Chen ◽  
Wei Wang ◽  
Wei Gao ◽  
Zhi-Hua Zhou
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Learning ◽  
Error Rate ◽  
Deep Neural Network ◽  
Deep Neural Networks ◽  
State Of The Art ◽  
Fine Tuning ◽  
Learning Methods ◽  
Model Initialization

Deep neural networks have witnessed great successes in various real applications, but it requires a large number of labeled data for training. In this paper, we propose tri-net, a deep neural network which is able to use massive unlabeled data to help learning with limited labeled data. We consider model initialization, diversity augmentation and pseudo-label editing simultaneously. In our work, we utilize output smearing to initialize modules, use fine-tuning on labeled data to augment diversity and eliminate unstable pseudo-labels to alleviate the influence of suspicious pseudo-labeled data. Experiments show that our method achieves the best performance in comparison with state-of-the-art semi-supervised deep learning methods. In particular, it achieves 8.30% error rate on CIFAR-10 by using only 4000 labeled examples.

Domain Randomization Using Deep Neural Networks for Estimating Positions of Bolts

2020 ◽  
Vol 20 (5) ◽  
Author(s):  
Ezra Ameperosa ◽  
Pranav A. Bhounsule
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Convolutional Neural Network ◽  
Deep Neural Networks ◽  
Single Object ◽  
Automated Method ◽  
Bounding Box ◽  
The Neural Network ◽  
Camera Position ◽  
Position And Orientation

Abstract Current manual practices of replacing bolts on structures are time-consuming and costly, especially because of numerous bolts. Thus, an automated method that can visually detect and localize bolt positions would be highly beneficial. We demonstrate the use of deep neural networks using domain randomization for detecting and localizing bolts on a workpiece. In contrast to previous approaches that require training on real images, the use of domain randomization enables all training in simulation. The key idea is to create a wide variety of computer-generated synthetic images by varying the texture, color, camera position and orientation, distractor objects, and noise, and train the neural network on these images such that the neural network is robust to scene variability and hence provides accurate results when deployed on real images. Using domain randomization, we train two neural networks, a faster regional convolutional neural network for detecting the bolt and placing a bounding box, and a regression convolutional neural network for estimating the x- and y-position of the bolts relative to the coordinates fixed to the workpiece. Our results indicate that in the best case, we can detect bolts with 85% accuracy and can predict 75% of bolts within 1.27 cm accuracy. The novelty of this work is in using domain randomization to detect and localize: (1) multiples of a single object and (2) small-sized objects (0.6 cm × 2.5 cm).

Sign in / Sign up

Export Citation Format

Share Document