Suspicious Behavior Detection with Temporal Feature Extraction and Time-Series Classification for Shoplifting Crime Prevention.

The rise in crime rates in many parts of the world, coupled with advancements in computer vision, has increased the need for automated crime detection services. To address this issue, we propose a new approach for detecting suspicious behavior as a means of preventing shoplifting. Existing methods are based on the use of convolutional neural networks that rely on extracting spatial features from pixel values. In contrast, our proposed method employs object detection based on YOLOv5 with Deep Sort to track people through a video, using the resulting bounding box coordinates as temporal features. The extracted temporal features are then modeled as a time-series classification problem. The proposed method was tested on the popular UCF Crime dataset, and benchmarked against the current state-of-the-art robust temporal feature magnitude (RTFM) method, which relies on the Inflated 3D ConvNet (I3D) preprocessing method. Our results demonstrate an impressive 8.45-fold increase in detection inference speed compared to the state-of-the-art RTFM, along with an F1 score of 92%,outperforming RTFM by 3%. Furthermore, our method achieved these results without requiring expensive data augmentation or image feature extraction.

Short term energy consumption forecasting using neural basis expansion analysis for interpretable time series.

Smart grids and smart homes are getting people's attention in the modern era of smart cities. The advancements of smart technologies and smart grids have created challenges related to energy efficiency and production according to the future demand of clients. Machine learning, specifically neural network-based methods, remained successful in energy consumption prediction, but still, there are gaps due to uncertainty in the data and limitations of the algorithms. Research published in the literature has used small datasets and profiles of primarily single users; therefore, models have difficulties when applied to large datasets with profiles of different customers. Thus, a smart grid environment requires a model that handles consumption data from thousands of customers. The proposed model enhances the newly introduced method of Neural Basis Expansion Analysis for interpretable Time Series (N-BEATS) with a big dataset of energy consumption of 169 customers. Further, to validate the results of the proposed model, a performance comparison has been carried out with the Long Short Term Memory (LSTM), Blocked LSTM, Gated Recurrent Units (GRU), Blocked GRU and Temporal Convolutional Network (TCN). The proposed interpretable model improves the prediction accuracy on the big dataset containing energy consumption profiles of multiple customers. Incorporating covariates into the model improved accuracy by learning past and future energy consumption patterns. Based on a large dataset, the proposed model performed better for daily, weekly, and monthly energy consumption predictions. The forecasting accuracy of the N-BEATS interpretable model for 1-day-ahead energy consumption with "day as covariates" remained better than the 1, 2, 3, and 4-week scenarios.

Interpretable deep learning for the prediction of ICU admission likelihood and mortality of COVID-19 patients.

The global healthcare system is being overburdened by an increasing number of COVID-19 patients. Physicians are having difficulty allocating resources and focusing their attention on high-risk patients, partly due to the difficulty in identifying high-risk patients early. COVID-19 hospitalizations require specialized treatment capabilities and can cause a burden on healthcare resources. Estimating future hospitalization of COVID-19 patients is, therefore, crucial to saving lives. In this paper, an interpretable deep learning model is developed to predict intensive care unit (ICU) admission and mortality of COVID-19 patients. The study comprised of patients from the Stony Brook University Hospital, with patient information such as demographics, comorbidities, symptoms, vital signs, and laboratory tests recorded. The top three predictors of ICU admission were ferritin, diarrhoea, and alamine aminotransferase, and the top predictors for mortality were COPD, ferritin, and myalgia. The proposed model predicted ICU admission with an AUC score of 88.3% and predicted mortality with an AUC score of 96.3%. The proposed model was evaluated against existing model in the literature which achieved an AUC of 72.8% in predicting ICU admission and achieved an AUC of 84.4% in predicting mortality. It can clearly be seen that the model proposed in this paper shows superiority over existing models. The proposed model has the potential to provide tools to frontline doctors to help classify patients in time-bound and resource-limited scenarios.

A Non-Invasive Interpretable Diagnosis of Melanoma Skin Cancer Using Deep Learning and Ensemble Stacking of Machine Learning Models.

A skin lesion is a portion of skin that observes abnormal growth compared to other areas of the skin. The ISIC 2018 lesion dataset has seven classes. A miniature dataset version of it is also available with only two classes: malignant and benign. Malignant tumors are tumors that are cancerous, and benign tumors are non-cancerous. Malignant tumors have the ability to multiply and spread throughout the body at a much faster rate. The early detection of the cancerous skin lesion is crucial for the survival of the patient. Deep learning models and machine learning models play an essential role in the detection of skin lesions. Still, due to image occlusions and imbalanced datasets, the accuracies have been compromised so far. In this paper, we introduce an interpretable method for the non-invasive diagnosis of melanoma skin cancer using deep learning and ensemble stacking of machine learning models. The dataset used to train the classifier models contains balanced images of benign and malignant skin moles. Hand-crafted features are used to train the base models (logistic regression, SVM, random forest, KNN, and gradient boosting machine) of machine learning. The prediction of these base models was used to train level one model stacking using cross-validation on the training set. Deep learning models (MobileNet, Xception, ResNet50, ResNet50V2, and DenseNet121) were used for transfer learning, and were already pre-trained on ImageNet data. The classifier was evaluated for each model. The deep learning models were then ensembled with different combinations of models and assessed. Furthermore, shapely adaptive explanations are used to construct an interpretability approach that generates heatmaps to identify the parts of an image that are most suggestive of the illness. This allows dermatologists to understand the results of our model in a way that makes sense to them. For evaluation, we calculated the accuracy, F1-score, Cohen's kappa, confusion matrix, and ROC curves and identified the best model for classifying skin lesions.

A hybrid multi objective cellular spotted hyena optimizer for wellbore trajectory optimization.

Cost and safety are critical factors in the oil and gas industry for optimizing wellbore trajectory, which is a constrained and nonlinear optimization problem. In this work, the wellbore trajectory is optimized using the true measured depth, well profile energy, and torque. Numerous metaheuristic algorithms were employed to optimize these objectives by tuning 17 constrained variables, with notable drawbacks including decreased exploitation/exploration capability, local optima trapping, non-uniform distribution of non-dominated solutions, and inability to track isolated minima. The purpose of this work is to propose a modified multi-objective cellular spotted hyena algorithm (MOCSHOPSO) for optimizing true measured depth, well profile energy, and torque. To overcome the aforementioned difficulties, the modification incorporates cellular automata (CA) and particle swarm optimization (PSO). By adding CA, the SHO's exploration phase is enhanced, and the SHO's hunting mechanisms are modified with PSO's velocity update property. Several geophysical and operational constraints have been utilized during trajectory optimization and data has been collected from the Gulf of Suez oil field. The proposed algorithm was compared with the standard methods (MOCPSO, MOSHO, MOCGWO) and observed significant improvements in terms of better distribution of non-dominated solutions, better-searching capability, a minimum number of isolated minima, and better Pareto optimal front. These significant improvements were validated by analysing the algorithms in terms of some statistical analysis, such as IGD, MS, SP, and ER. The proposed algorithm has obtained the lowest values in IGD, SP and ER, on the other side highest values in MS. Finally, an adaptive neighbourhood mechanism has been proposed which showed better performance than the fixed neighbourhood topology such as L5, L9, C9, C13, C21, and C25. Hopefully, this newly proposed modified algorithm will pave the way for better wellbore trajectory optimization.