scholarly journals Explaining Deep Learning Models for Tabular Data Using Layer-Wise Relevance Propagation

2021 ◽  
Vol 12 (1) ◽  
pp. 136
Author(s):  
Ihsan Ullah ◽  
Andre Rios ◽  
Vaibhav Gala ◽  
Susan Mckeever
Keyword(s):  
Deep Learning ◽  
Credit Card ◽  
Model Performance ◽  
Feature Subset Selection ◽  
Mixed Data ◽  
Computational Time ◽  
Feature Subset ◽  
Learning Models ◽  
Customer Churn ◽  
Interpretable Model

Trust and credibility in machine learning models are bolstered by the ability of a model to explain its decisions. While explainability of deep learning models is a well-known challenge, a further challenge is clarity of the explanation itself for relevant stakeholders of the model. Layer-wise Relevance Propagation (LRP), an established explainability technique developed for deep models in computer vision, provides intuitive human-readable heat maps of input images. We present the novel application of LRP with tabular datasets containing mixed data (categorical and numerical) using a deep neural network (1D-CNN), for Credit Card Fraud detection and Telecom Customer Churn prediction use cases. We show how LRP is more effective than traditional explainability concepts of Local Interpretable Model-agnostic Explanations (LIME) and Shapley Additive Explanations (SHAP) for explainability. This effectiveness is both local to a sample level and holistic over the whole testing set. We also discuss the significant computational time advantage of LRP (1–2 s) over LIME (22 s) and SHAP (108 s) on the same laptop, and thus its potential for real time application scenarios. In addition, our validation of LRP has highlighted features for enhancing model performance, thus opening up a new area of research of using XAI as an approach for feature subset selection.

scholarly journals Optimal Segmentation Scale Parameter, Feature Subset and Classification Algorithm for Geographic Object-Based Crop Recognition Using Multisource Satellite Imagery

2019 ◽  
Vol 11 (5) ◽  
pp. 514 ◽  
Author(s):  
Lingbo Yang ◽  
Lamin Mansaray ◽  
Jingfeng Huang ◽  
Limin Wang
Keyword(s):  
Satellite Imagery ◽  
Scale Parameter ◽  
Selection Method ◽  
Classification Algorithm ◽  
Feature Subset Selection ◽  
Computational Time ◽  
Feature Subset ◽  
Object Based ◽  
Geographic Object ◽  
Crop Classification

Geographic object-based image analysis (GEOBIA) has been widely used in the remote sensing of agricultural crops. However, issues related to image segmentation, data redundancy and performance of different classification algorithms with GEOBIA have not been properly addressed in previous studies, thereby compromising the accuracy of subsequent thematic products. It is in this regard that the current study investigates the optimal scale parameter (SP) in multi-resolution segmentation, feature subset, and classification algorithm for use in GEOBIA based on multisource satellite imagery. For this purpose, a novel supervised optimal SP selection method was proposed based on information gain ratio, and was then compared with a preexisting unsupervised optimal SP selection method. Additionally, the recursive feature elimination (RFE) and enhanced RFE (EnRFE) algorithms were modified to generate an improved EnRFE (iEnRFE) algorithm, which was then compared with its precursors in the selection of optimal classification features. Based on the above, random forest (RF), gradient boosting decision tree (GBDT) and support vector machine (SVM) were applied to segmented objects for crop classification. The results indicated that the supervised optimal SP selection method is more suitable for application in heterogeneous land cover, whereas the unsupervised method proved more efficient as it does not require reference segmentation objects. The proposed iEnRFE method outperformed the preexisting EnRFE and RFE methods in optimal feature subset selection as it recorded the highest accuracy and less processing time. The RF, GBDT, and SVM algorithms achieved overall classification accuracies of 91.8%, 92.4%, and 90.5%, respectively. GBDT and RF recorded higher classification accuracies and utilized much less computational time than SVM and are, therefore, considered more suitable for crop classification requiring large numbers of image features. These results have shown that the proposed object-based crop classification scheme could provide a valuable reference for relevant applications of GEOBIA in crop recognition using multisource satellite imagery.

scholarly journals Improving generalization of deep learning models for diagnostic pathology by increasing variability in training data: experiments on osteosarcoma subtypes

2020 ◽  
Author(s):  
Haiming Tang ◽  
Nanfei Sun ◽  
Steven Shen
Keyword(s):  
Deep Learning ◽  
Model Performance ◽  
High Variability ◽  
Training Data ◽  
Classification Model ◽  
Training Dataset ◽  
Learning Models ◽  
Diagnostic Pathology ◽  
Model Generalization ◽  
Histopathological Images

Artificial intelligence (AI) has an emerging progress in diagnostic pathology. A large number of studies of applying deep learning models to histopathological images have been published in recent years. While many studies claim high accuracies, they may fall into the pitfalls of overfitting and lack of generalization due to the high variability of the histopathological images. We use the example of Osteosarcoma to illustrate the pitfalls and how the addition of model input variability can help improve model performance. We use the publicly available osteosarcoma dataset to retrain a previously published classification model for osteosarcoma. We partition the same set of images into the training and testing datasets differently than the original study: the test dataset consists of images from one patient while the training dataset consists images of all other patients. The performance of the model on the test set using the new partition schema declines dramatically, indicating a lack of model generalization and overfitting.We also show the influence of training data variability on model performance by collecting a minimal dataset of 10 osteosarcoma subtypes as well as benign tissues and benign bone tumors of differentiation. We show the additions of more and more subtypes into the training data step by step under the same model schema yield a series of coherent models with increasing performances. In conclusion, we bring forward data preprocessing and collection tactics for histopathological images of high variability to avoid the pitfalls of overfitting and build deep learning models of higher generalization abilities.

scholarly journals Effect of Sequence Padding on the Performance of Protein-Based Deep Learning Models

2020 ◽  
Author(s):  
Angela Lopez-del Rio ◽  
Maria Martin ◽  
Alexandre Perera-Lluna ◽  
Rabie Saidi
Keyword(s):  
Deep Learning ◽  
Amino Acid ◽  
Enzyme Commission ◽  
Model Performance ◽  
Enzyme Commission Number ◽  
Amino Acid Sequences ◽  
Learning Models ◽  
Zero Padding ◽  
The One ◽  
The Impact

Abstract Background The use of raw amino acid sequences as input for protein-based deep learning models has gained popularity in recent years. This scheme obliges to manage proteins with different lengths, while deep learning models require same-shape input. To accomplish this, zeros are usually added to each sequence up to a established common length in a process called zero-padding. However, the effect of different padding strategies on model performance and data structure is yet unknown. Results We analysed the impact of different ways of padding the amino acid sequences in a hierarchical Enzyme Commission number prediction problem. Our results show that padding has an effect on model performance even when there are convolutional layers implied. We propose and implement four novel types of padding the amino acid sequences. Conclusions The present study highlights the relevance of the step of padding the one-hot encoded amino acid sequences when building deep learning-based models for Enzyme Commission number prediction. The fact that this has an effect on model performance should raise awareness on the need of justifying the details of this step on future works. The code of this analysis is available at https://github.com/b2slab/padding_benchmark.

scholarly journals Emotional Classification of Acoustic Information With Optimal Feature Subset Selection Methods

2018 ◽  
Vol 7 (2.32) ◽  
pp. 39
Author(s):  
Dr Swarna Kuchibhotla ◽  
Mr Niranjan M.S.R
Keyword(s):  
High Efficiency ◽  
Subset Selection ◽  
Feature Subset Selection ◽  
Computational Time ◽  
Feature Subset ◽  
Selection Methods ◽  
Forward Selection ◽  
Optimal Feature Subset ◽  
Optimal Feature

This paper mainly focuses on classification of various Acoustic emotional corpora with frequency domain features using feature subset selection methods. The emotional speech samples are classified into neutral,  happy, fear , anger,  disgust and sad  states by using properties of statistics  of spectral features estimated from Berlin and Spanish emotional utterances. The Sequential Forward Selection(SFS) and Sequential Floating Forward Selection(SFFS)feature subset selection algorithms are  for extracting more informative features. The number of speech emotional samples available for training is smaller than that of the number of features extracted from the speech sample in both Berlin and Spanish corpora which is called curse of dimensionality. Because of this  feature vector of high dimensionality the efficiency of the classifier decreases and at the same time the computational time also increases. For additional  improvement in the efficiency of the classifier  a subset of  features which are optimal is needed and is obtained by using feature subset selection methods. This will enhances the performance of the system with high efficiency and lower computation time. The classifier used in this work is the standard K Nearest Neighbour (KNN) Classifier. Experimental evaluation   proved  that the performance of the classifier is enhanced with SFFS because it vanishes the nesting effect suffered by SFS. The results also showed that an optimal feature subset is a better choice for classification rather than full feature set.  

scholarly journals Optimising Deep Learning at the Edge for Accurate Hourly Air Quality Prediction

Sensors ◽  
2021 ◽  
Vol 21 (4) ◽  
pp. 1064
Author(s):  
I Nyoman Kusuma Wardana ◽  
Julian W. Gardner ◽  
Suhaib A. Fahmy
Keyword(s):  
Deep Learning ◽  
Air Quality ◽  
Short Term Memory ◽  
Model Performance ◽  
Learning Model ◽  
Sensor Data ◽  
Raspberry Pi ◽  
Learning Models ◽  
Proposed Model ◽  
Deep Learning Model

Accurate air quality monitoring requires processing of multi-dimensional, multi-location sensor data, which has previously been considered in centralised machine learning models. These are often unsuitable for resource-constrained edge devices. In this article, we address this challenge by: (1) designing a novel hybrid deep learning model for hourly PM2.5 pollutant prediction; (2) optimising the obtained model for edge devices; and (3) examining model performance running on the edge devices in terms of both accuracy and latency. The hybrid deep learning model in this work comprises a 1D Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) to predict hourly PM2.5 concentration. The results show that our proposed model outperforms other deep learning models, evaluated by calculating RMSE and MAE errors. The proposed model was optimised for edge devices, the Raspberry Pi 3 Model B+ (RPi3B+) and Raspberry Pi 4 Model B (RPi4B). This optimised model reduced file size to a quarter of the original, with further size reduction achieved by implementing different post-training quantisation. In total, 8272 hourly samples were continuously fed to the edge device, with the RPi4B executing the model twice as fast as the RPi3B+ in all quantisation modes. Full-integer quantisation produced the lowest execution time, with latencies of 2.19 s and 4.73 s for RPi4B and RPi3B+, respectively.

scholarly journals Generalizability of deep learning models for dental image analysis

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Joachim Krois ◽  
Anselmo Garcia Cantu ◽  
Akhilanand Chaurasia ◽  
Ranjitkumar Patil ◽  
Prabhat Kumar Chaudhari ◽  
...  
Keyword(s):  
Deep Learning ◽  
Root Canal ◽  
Model Performance ◽  
Image Data ◽  
Dental Status ◽  
Learning Models ◽  
Panoramic Radiographs ◽  
Image Characteristics ◽  
Pixel Value ◽  
The Impact

AbstractWe assessed the generalizability of deep learning models and how to improve it. Our exemplary use-case was the detection of apical lesions on panoramic radiographs. We employed two datasets of panoramic radiographs from two centers, one in Germany (Charité, Berlin, n = 650) and one in India (KGMU, Lucknow, n = 650): First, U-Net type models were trained on images from Charité (n = 500) and assessed on test sets from Charité and KGMU (each n = 150). Second, the relevance of image characteristics was explored using pixel-value transformations, aligning the image characteristics in the datasets. Third, cross-center training effects on generalizability were evaluated by stepwise replacing Charite with KGMU images. Last, we assessed the impact of the dental status (presence of root-canal fillings or restorations). Models trained only on Charité images showed a (mean ± SD) F1-score of 54.1 ± 0.8% on Charité and 32.7 ± 0.8% on KGMU data (p < 0.001/t-test). Alignment of image data characteristics between the centers did not improve generalizability. However, by gradually increasing the fraction of KGMU images in the training set (from 0 to 100%) the F1-score on KGMU images improved (46.1 ± 0.9%) at a moderate decrease on Charité images (50.9 ± 0.9%, p < 0.01). Model performance was good on KGMU images showing root-canal fillings and/or restorations, but much lower on KGMU images without root-canal fillings and/or restorations. Our deep learning models were not generalizable across centers. Cross-center training improved generalizability. Noteworthy, the dental status, but not image characteristics were relevant. Understanding the reasons behind limits in generalizability helps to mitigate generalizability problems.

scholarly journals Comparison of Deep Learning Models and Various Text Pre-Processing Techniques for the Toxic Comments Classification

2020 ◽  
Vol 10 (23) ◽  
pp. 8631
Author(s):  
Viera Maslej-Krešňáková ◽  
Martin Sarnovský ◽  
Peter Butka ◽  
Kristína Machová
Keyword(s):  
Deep Learning ◽  
Language Processing ◽  
Social Behaviour ◽  
Model Performance ◽  
Language Models ◽  
Learning Models ◽  
Processing Methods ◽  
Detection And Identification ◽  
Processing Techniques ◽  
Modern Machine

The emergence of anti-social behaviour in online environments presents a serious issue in today’s society. Automatic detection and identification of such behaviour are becoming increasingly important. Modern machine learning and natural language processing methods can provide effective tools to detect different types of anti-social behaviour from the pieces of text. In this work, we present a comparison of various deep learning models used to identify the toxic comments in the Internet discussions. Our main goal was to explore the effect of the data preparation on the model performance. As we worked with the assumption that the use of traditional pre-processing methods may lead to the loss of characteristic traits, specific for toxic content, we compared several popular deep learning and transformer language models. We aimed to analyze the influence of different pre-processing techniques and text representations including standard TF-IDF, pre-trained word embeddings and also explored currently popular transformer models. Experiments were performed on the dataset from the Kaggle Toxic Comment Classification competition, and the best performing model was compared with the similar approaches using standard metrics used in data analysis.

scholarly journals A Novel Bio-Inspired Deep Learning Approach for Liver Cancer Diagnosis

Information ◽  
2020 ◽  
Vol 11 (2) ◽  
pp. 80 ◽  
Author(s):  
Rania M. Ghoniem
Keyword(s):  
Deep Learning ◽  
Liver Cancer ◽  
State Of The Art ◽  
The State ◽  
Convergence Time ◽  
Computational Time ◽  
Learning Approach ◽  
Liver Lesions ◽  
Learning Models ◽  
Abc Algorithm

Current research on computer-aided diagnosis (CAD) of liver cancer is based on traditional feature engineering methods, which have several drawbacks including redundant features and high computational cost. Recent deep learning models overcome these problems by implicitly capturing intricate structures from large-scale medical image data. However, they are still affected by network hyperparameters and topology. Hence, the state of the art in this area can be further optimized by integrating bio-inspired concepts into deep learning models. This work proposes a novel bio-inspired deep learning approach for optimizing predictive results of liver cancer. This approach contributes to the literature in two ways. Firstly, a novel hybrid segmentation algorithm is proposed to extract liver lesions from computed tomography (CT) images using SegNet network, UNet network, and artificial bee colony optimization (ABC), namely, SegNet-UNet-ABC. This algorithm uses the SegNet for separating liver from the abdominal CT scan, then the UNet is used to extract lesions from the liver. In parallel, the ABC algorithm is hybridized with each network to tune its hyperparameters, as they highly affect the segmentation performance. Secondly, a hybrid algorithm of the LeNet-5 model and ABC algorithm, namely, LeNet-5/ABC, is proposed as feature extractor and classifier of liver lesions. The LeNet-5/ABC algorithm uses the ABC to select the optimal topology for constructing the LeNet-5 network, as network structure affects learning time and classification accuracy. For assessing performance of the two proposed algorithms, comparisons have been made to the state-of-the-art algorithms on liver lesion segmentation and classification. The results reveal that the SegNet-UNet-ABC is superior to other compared algorithms regarding Jaccard index, Dice index, correlation coefficient, and convergence time. Moreover, the LeNet-5/ABC algorithm outperforms other algorithms regarding specificity, F1-score, accuracy, and computational time.

scholarly journals Deep Sparse Learning for Automatic Modulation Classification Using Recurrent Neural Networks

Sensors ◽  
2021 ◽  
Vol 21 (19) ◽  
pp. 6410
Author(s):  
Ke Zang ◽  
Wenqi Wu ◽  
Wei Luo
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Recurrent Neural Networks ◽  
Deep Neural Networks ◽  
Learning Algorithm ◽  
Model Performance ◽  
Sparse Learning ◽  
Modulation Classification ◽  
Learning Models ◽  
Automatic Modulation Classification

Deep learning models, especially recurrent neural networks (RNNs), have been successfully applied to automatic modulation classification (AMC) problems recently. However, deep neural networks are usually overparameterized, i.e., most of the connections between neurons are redundant. The large model size hinders the deployment of deep neural networks in applications such as Internet-of-Things (IoT) networks. Therefore, reducing parameters without compromising the network performance via sparse learning is often desirable since it can alleviates the computational and storage burdens of deep learning models. In this paper, we propose a sparse learning algorithm that can directly train a sparsely connected neural network based on the statistics of weight magnitude and gradient momentum. We first used the MNIST and CIFAR10 datasets to demonstrate the effectiveness of this method. Subsequently, we applied it to RNNs with different pruning strategies on recurrent and non-recurrent connections for AMC problems. Experimental results demonstrated that the proposed method can effectively reduce the parameters of the neural networks while maintaining model performance. Moreover, we show that appropriate sparsity can further improve network generalization ability.

Sign in / Sign up

Export Citation Format

Share Document