Visualization of Speech Perception Analysis via Phoneme Alignment: A Pilot Study

Objective: Speech tests assess the ability of people with hearing loss to comprehend speech with a hearing aid or cochlear implant. The tests are usually at the word or sentence level. However, few tests analyze errors at the phoneme level. So, there is a need for an automated program to visualize in real time the accuracy of phonemes in these tests.Method: The program reads in stimulus-response pairs and obtains their phonemic representations from an open-source digital pronouncing dictionary. The stimulus phonemes are aligned with the response phonemes via a modification of the Levenshtein Minimum Edit Distance algorithm. Alignment is achieved via dynamic programming with modified costs based on phonological features for insertion, deletions and substitutions. The accuracy for each phoneme is based on the F1-score. Accuracy is visualized with respect to place and manner (consonants) or height (vowels). Confusion matrices for the phonemes are used in an information transfer analysis of ten phonological features. A histogram of the information transfer for the features over a frequency-like range is presented as a phonemegram.Results: The program was applied to two datasets. One consisted of test data at the sentence and word levels. Stimulus-response sentence pairs from six volunteers with different degrees of hearing loss and modes of amplification were analyzed. Four volunteers listened to sentences from a mobile auditory training app while two listened to sentences from a clinical speech test. Stimulus-response word pairs from three lists were also analyzed. The other dataset consisted of published stimulus-response pairs from experiments of 31 participants with cochlear implants listening to 400 Basic English Lexicon sentences via different talkers at four different SNR levels. In all cases, visualization was obtained in real time. Analysis of 12,400 actual and random pairs showed that the program was robust to the nature of the pairs.Conclusion: It is possible to automate the alignment of phonemes extracted from stimulus-response pairs from speech tests in real time. The alignment then makes it possible to visualize the accuracy of responses via phonological features in two ways. Such visualization of phoneme alignment and accuracy could aid clinicians and scientists.

Techniques to Deal with Off-Diagonal Elements in Confusion Matrices

Confusion matrices are numerical structures that deal with the distribution of errors between different classes or categories in a classification process. From a quality perspective, it is of interest to know if the confusion between the true class A and the class labelled as B is not the same as the confusion between the true class B and the class labelled as A. Otherwise, a problem with the classifier, or of identifiability between classes, may exist. In this paper two statistical methods are considered to deal with this issue. Both of them focus on the study of the off-diagonal cells in confusion matrices. First, McNemar-type tests to test the marginal homogeneity are considered, which must be followed from a one versus all study for every pair of categories. Second, a Bayesian proposal based on the Dirichlet distribution is introduced. This allows us to assess the probabilities of misclassification in a confusion matrix. Three applications, including a set of omic data, have been carried out by using the software R.

Predicting stroke risk by Migraine using AI

Stroke is a blood clot or bleeds in the brain, which can make permanent damage that has an effect on mobility, cognition, sight or communication. It is the second leading cause of death worldwide and one of the most life- threatening diseases for persons above 65 years. It damages the brain like “heart attack” which damages the heart. Every 4 minutes someone dies of stroke, but up to 80% of stroke can be prevented if we can identify or predict the occurrence of stroke in its early stage. In this paper, I used different types of machine learning algorithms for stroke prediction on the Healthcare Dataset Stroke data. Four types of machine learning classification algorithms were applied; Linear Regression, Confusion matrices, Random Forest Classifier, and Logistic Regression were used to build the stroke prediction model. Support, Precision, Recall, and F1-score were used to calculate performance measures of machine learning models. The results showed that Random Forest Classifier has achieved the best accuracy at 94 % [1].

Comparative Study of Popular Deep Learning Models for Machining Roughness Classification Using Sound and Force Signals

This study compared popular Deep Learning (DL) architectures to classify machining surface roughness using sound and force data. The DL architectures considered in this study include Multi-Layer Perceptron (MLP), Convolution Neural Network (CNN), Long Short-Term Memory (LSTM), and transformer. The classification was performed on the sound and force data generated during machining aluminum sheets for different levels of spindle speed, feed rate, depth of cut, and end-mill diameter, and it was trained on 30 s machining data (10–40 s) of the machining experiments. Since a raw audio waveform is seldom used in DL models, Mel-Spectrogram and Mel Frequency Cepstral Coefficients (MFCCs) audio feature extraction techniques were used in the DL models. The results of DL models were compared for the training–validation accuracy, training epochs, and training parameters of each model. Although the roughness classification by all the DL models was satisfactory (except for CNN with Mel-Spectrogram), the transformer-based modes had the highest training (>96%) and validation accuracies (≈90%). The CNN model with Mel-Spectrogram exhibited the worst training and inference accuracy, which is influenced by limited training data. Confusion matrices were plotted to observe the classification accuracy visually. The confusion matrices showed that the transformer model trained on Mel-Spectrogram and the transformer model trained on MFCCs correctly predicted 366 (or 91.5%) and 371 (or 92.7%) out of 400 test samples. This study also highlights the suitability and superiority of the transformer model for time series sound and force data and over other DL models.

Prediction of Inpatient Pressure Ulcers Based on Routine Healthcare Data Using Machine Learning Methodology

Despite the relevance of pressure ulcers (PU) in inpatient care, the predictive power and role of care-related risk factors (e.g. surgical anesthesia) remain unclear. We investigated the predictability of PU incidence and its association with multiple care variables. We included all somatic cases between 2014 and 2018 with length of stay ≥2 days in a German university hospital. For regression analyses and prediction we used Bayesian Additive Regression Trees (BART) as nonparametric modeling approach. To assess predictive accuracy, we compared BART and logistic regression (LR) using area under the curve (AUC) and confusion matrices. The analysis of 149,006 cases revealed high predictive variable importance and associations between incident PU and intensive care with ventilation, age, surgical anesthesia (≥1 hour) and number of care-involved wards. Despite high AUCs (LR: 0.89; BART: 0.9), the confusion matrices showed a higher number of false negative (LR: 816; BART: 826) than true positive (LR: 138; BART: 68) predictions. In summary, particularly intensive care with ventilation, age, anesthesia and number of care-involved wards were associated with incident PU. Using surgical anesthesia as a proxy for immobility, our results suggest hourly repositioning. High rates of false negative predictions indicate a general challenge in the predictability of PU.

BIKED: A Dataset for Computational Bicycle Design with Machine Learning Benchmarks

In this paper, we present “BIKED,” a dataset comprised of 4500 individually designed bicycle models sourced from hundreds of designers. We expect BIKED to enable a variety of data-driven design applications for bicycles and support the development of data-driven design methods. The dataset is comprised of a variety of design information including assembly images, component images, numerical design parameters, and class labels. In this paper, we first discuss the processing of the dataset, then highlight some prominent research questions that BIKED can help address. Of these questions, we further explore the following in detail: 1) How can we explore, understand, and visualize the current design space of bicycles and utilize this information? We apply unsupervised embedding methods to study the design space and identify key takeaways from this analysis. 2) When designing bikes using algorithms, under what conditions can machines understand the design of a given bike? We train a multitude of classifiers to understand designs, then examine the behavior of these classifiers through confusion matrices and permutation-based interpretability analysis. 3) Can machines learn to synthesize new bicycle designs by studying existing ones? We test Variational Autoencoders on random generation, interpolation, and extrapolation tasks after training on BIKED data. The dataset and code are available at http://decode.mit.edu/projects/biked/

Predictive risk model of livestock rabies occurrence in Peru

Livestock rabies is endemic in Peru. Hence, its persistence and annual dissemination represent an important economic impact, especially for impoverished farming communities. The disease is mostly transmitted by the hematophagous bat Desmodus rotundus. The present study aimed to adapt an existing predictive model of the occurrence of livestock rabies to Peru, in which the risk of rabies transmission from bats to livestock was estimated using decision-tree models of receptivity and vulnerability. Official rabies surveillance data between 2010 and 2015 were used along with possible risk factors, such as livestock biomass, environmental changes, and geomorphological characteristics. Several scenarios were established to evaluate the prediction of the occurrence of livestock rabies cases by determining more than one cut-off point of the receptivity variables. During the study period, the precision of the model was estimated through the sensitivity (39.46%) and specificity (98.64%) by using confusion matrices. Targeting control efforts, especially in districts with a high estimated risk, could represent the prevention of a significant proportion of livestock rabies cases, which would optimize the human and economic resources of the Peruvian surveillance service. However, the quality of data produced by the surveillance should be improved not only to obtain higher model precision but also to allow the adequate planning of control actions.

Detection of COVID-19 Using Transfer Learning and Grad-CAM Visualization on Indigenously Collected X-ray Dataset

The COVID-19 outbreak began in December 2019 and has dreadfully affected our lives since then. More than three million lives have been engulfed by this newest member of the corona virus family. With the emergence of continuously mutating variants of this virus, it is still indispensable to successfully diagnose the virus at early stages. Although the primary technique for the diagnosis is the PCR test, the non-contact methods utilizing the chest radiographs and CT scans are always preferred. Artificial intelligence, in this regard, plays an essential role in the early and accurate detection of COVID-19 using pulmonary images. In this research, a transfer learning technique with fine tuning was utilized for the detection and classification of COVID-19. Four pre-trained models i.e., VGG16, DenseNet-121, ResNet-50, and MobileNet were used. The aforementioned deep neural networks were trained using the dataset (available on Kaggle) of 7232 (COVID-19 and normal) chest X-ray images. An indigenous dataset of 450 chest X-ray images of Pakistani patients was collected and used for testing and prediction purposes. Various important parameters, e.g., recall, specificity, F1-score, precision, loss graphs, and confusion matrices were calculated to validate the accuracy of the models. The achieved accuracies of VGG16, ResNet-50, DenseNet-121, and MobileNet are 83.27%, 92.48%, 96.49%, and 96.48%, respectively. In order to display feature maps that depict the decomposition process of an input image into various filters, a visualization of the intermediate activations is performed. Finally, the Grad-CAM technique was applied to create class-specific heatmap images in order to highlight the features extracted in the X-ray images. Various optimizers were used for error minimization purposes. DenseNet-121 outperformed the other three models in terms of both accuracy and prediction.

Classification of Cardiomyopathies from MR Cine Images Using Convolutional Neural Network with Transfer Learning

The automatic classification of various types of cardiomyopathies is desirable but has never been performed using a convolutional neural network (CNN). The purpose of this study was to evaluate currently available CNN models to classify cine magnetic resonance (cine-MR) images of cardiomyopathies. Method: Diastolic and systolic frames of 1200 cine-MR sequences of three categories of subjects (395 normal, 411 hypertrophic cardiomyopathy, and 394 dilated cardiomyopathy) were selected, preprocessed, and labeled. Pretrained, fine-tuned deep learning models (VGG) were used for image classification (sixfold cross-validation and double split testing with hold-out data). The heat activation map algorithm (Grad-CAM) was applied to reveal salient pixel areas leading to the classification. Results: The diastolic–systolic dual-input concatenated VGG model cross-validation accuracy was 0.982 ± 0.009. Summed confusion matrices showed that, for the 1200 inputs, the VGG model led to 22 errors. The classification of a 227-input validation group, carried out by an experienced radiologist and cardiologist, led to a similar number of discrepancies. The image preparation process led to 5% accuracy improvement as compared to nonprepared images. Grad-CAM heat activation maps showed that most misclassifications occurred when extracardiac location caught the attention of the network. Conclusions: CNN networks are very well suited and are 98% accurate for the classification of cardiomyopathies, regardless of the imaging plane, when both diastolic and systolic frames are incorporated. Misclassification is in the same range as inter-observer discrepancies in experienced human readers.

The Comparison of Forward and Backward Neural Network Model – A Study on the Prediction of Student Grade

A neural network model can be used effectively in predicting training accuracy using machine learning. Based on the comparison of forward and backward neural networks, coded to communicate their output in the requisite manner using machine language is the basis of the present study. With the help of students' background information, to predict the Grade Point Average (GPA) of 580 engineering students based on various parameters, including mental health. The study is based on the Boruta algorithm and the random forest methods for data preparation in the matrices (12 * 2 = 24) of single-layered, multiple-layers, and forward and reverse algorithms adopted to test the prediction and accuracy of the grade point average by analyzing histograms, confusion matrices, and regression analysis. This study suggests the best model for predictions with the help of artificial neuron network that has roughly half the number of single layers and with three hidden layers.