Rapidly expanding domains such as the Internet of Things require sophisticated approaches to securing interconnected devices against cyber threats. The following study intends to fill in a crucial gap in the state of effective intrusion detection systems for the Internet of Things based on a comparison and analysis of various hyperparameter optimization approaches to improve existing and future detection systems. In other words, our main goal was to investigate and compare various hyperparameter optimization strategies to find and assess the most effective way to improve the performance of deep learning -based IDS. Our methodology was comprised of the following comparative optimization analysis used to compare a hybrid optimization approach against stand-alone implementation of Harmony Search and Bayesian Optimization. The analysis was done quantitatively based on IDS trained and tested on simulated Internet of Things network data, and IDS performance was evaluated by the following metrics : accuracy, precision, recall, and F1 score. The comparison of results showed that the hybrid optimization demonstrated the best performance indicators in terms of accuracy at 99.74%, precision at 99.7%, recall at 99.72%, and F1 score at 99.71%. The results of the study confirm the efficiency of implementing multiple optimization approaches and reveal the potential effectiveness of such combination for effective hyperparameter optimization of deep learning -based IDS in the Internet of Things environment.

2Introduction: Heart failure is a clinical syndrome resulting from structural or functional abnormalities of the heart, leading to reduced cardiac output or increased intracardiac pressure. When combined with cardiogenic shock, it becomes an emergency condition with a high mortality rate, necessitating immediate diagnosis and treatment. Accurate prediction of 30-day mortality in these patients is vital for timely care and patient survival. This study aimed to optimize the Random Forest algorithm by adjusting hyperparameters to more accurately predict 30-day mortality in heart failure patients with cardiogenic shock. Method: In this research, data from 201 cardiac patients aged over 18 years who experienced cardiogenic shock at Rouhani Hospital in Babol in 2020, were used. Thirty-four selected features such as age, history of cardiac surgery, pH, lactate levels, diabetes, etc., were examined, and their one-month mortality was tracked through telephone follow-ups. Results: The results showed that increasing age (above 57 years), decreasing pH (below 7. 3), and elevating lactate levels (above 2) significantly increased the risk of 30-day mortality. By optimizing the hyperparameters of the Random Forest algorithm (ntree=1000 and mtry=14), prediction accuracy improved from 66. 0% to 71. 8%. Conclusion: This study demonstrates that the accuracy of the Random Forest algorithm depends on its input hyperparameters and that optimizing these parameters can lead to a more precise prediction of mortality in heart failure patients with cardiogenic shock. With appropriate optimization, this algorithm can serve as an effective tool for the early detection of high-risk patients and timely provision.

Parkinson’s disease (PD) is a neurodegenerative disorder that progressively worsens with age, particularly affecting the elderly. Symptoms of PD include visual hallucinations, depression, autonomic dysfunction, and motor difficulties. Conventional diagnostic methods often rely on subjective interpretations of movement, which can be subtle and challenging to assess accurately, potentially leading to misdiagnoses. However, recent studies indicate that over 90% of individuals with PD exhibit vocal abnormalities at the onset of the disease. Machine learning (ML) techniques have shown promise in addressing these diagnostic challenges due to their higher efficiency and reduced error rates in analyzing complex, high-dimensional datasets, particularly those derived from speech signals. This study investigates 12 machine learning models—logistic regression (LR), support vector machine (SVM, linear/RBF), K-nearest neighbor (KNN), Naïve bayes (NB), decision tree (DT), random forest (RF), extra trees (ET), gradient boosting (GbBoost), extreme gradient boosting (XgBoost), adaboost, and multi-layer perceptron (MLP)—to develop a robust ML model capable of reliably identifying PD cases. The analysis utilized a PD voice dataset comprising 756 acoustic samples from 252 participants, including 188 individuals with PD and 64 healthy controls. The dataset included 130 male and 122 female subjects, with age ranges of 33 - 87 years and 41 - 82 years, respectively. To enhance model performance, the GridSearchCV method was employed for hyperparameter tuning, alongside recursive feature elimination (RFE) and minimum redundancy maximum relevance (mRMR) feature selection techniques. Among the 12 ML models evaluated, the RF model with the RFE-generated feature subset (RFE-50) emerged as the top performer. It achieved an accuracy of 96.46%, a recall of 0.96, a precision of 0.97, an F1-score of 0.96, and an AUC score of 0.998, marking the highest performance metrics reported for this dataset in recent studies.

In some countries, people commonly consume hazelnuts in their shells to extend shelf life or due to technological limitations. Therefore, open-shell hazelnuts are more marketable. At the semi-industrial scale, open-shell and closed-shell hazelnuts are currently separated from each other through visual inspection. This study aims to develop a new algorithm to separate open-shell hazelnuts from cracked or closed-shell hazelnuts. In the first approach, dimension reduction techniques such as Sequential Forward Feature Selection (SFFS) and Principal Component Analysis (PCA) were used to select or extract a combination of color, texture, and grayscale features for the model’s input. In the second approach, individual features were used directly as inputs. In this study, three famous machine learning models, including Support Vector Machine (SVM), K-nearest neighbors (KNN), and Multi-Layer Perceptron (MLP) were employed. The results indicated that the SFFS method had a greater effect on improving the performance of the models than the PCA method. However, there was no significant difference between the performance of the models developed with combined features (98.00%) and that of the models using individual features (98.67%). The overall results of this study indicated that the MLP model, with one hidden layer, a dropout of 0.3, and 10 neurons using Histogram of Oriented Gradients (HOG) features as input, is a good choice for classifying hazelnuts into two classes of open-shell and closed-shell.

Background: P300 signal detection is an essential problem in many fields of Brain-Computer Interface (BCI) systems. Although deep neural networks have almost ubiquitously used in P300 detection, in such networks, increasing the number of dimensions leads to growth ratio of saddle points to local minimums. This phenomenon results in slow convergence in deep neural network. hyperparameter tuning is one of the approaches in deep learning, which leads to fast convergence because of its ability to find better local minimums. In this paper, a new adaptive hyperparameter tuning method is proposed to improve training of Convolutional Neural Networks (CNNs). Methods: The aim of this paper is to introduce a novel method to improve the performance of deep neural networks in P300 signal detection. To reach this purpose, the proposed method transferred the non-convex error function of CNN) into Lagranging paradigm, then, Newton and dual active set techniques are utilized for hyperparameter tuning in order to minimize error of objective function in high dimensional space of CNN. Results: The proposed method was implemented on MATLAB 2017 package and its performance was evaluated on dataset of Ecole Polytechnique Fé dé rale de Lausanne (EPFL) BCI group. The obtained results depicted that the proposed method detected the P300 signals with 95. 34% classification accuracy in parallel with high True Positive Rate (i. e., 92. 9%) and low False Positive Rate (i. e., 0. 77%). Conclusions: To estimate the performance of the proposed method, the achieved results were compared with the results of Naive hyperparameter (NHP) tuning method. The comparisons depicted the superiority of the proposed method against its alternative, in such way that the best accuracy by using the proposed method was 6. 44%, better than the accuracy of the alternative method.

This study investigates the efficiency of CNN-DRNN hybrid classifiers in detecting sleep apnea using electrocardiogram (ECG) signals. Various CNN models were evaluated, including AlexNet, VGG16, VGG19, and ZFNet, along with DRNN models such as LSTM, GRU, and BiLSTM. These models were compared with and without the application of swarm intelligence optimizers, namely the Honey Badger Algorithm (HBA) and Grey Wolf Optimizer (GWO), for optimizing hyperparameter values. The results demonstrated that the AlexNet-GRU hybrid model achieved the best performance after applying both optimizers, with an accuracy of 95%, a detection rate of 97.61%, and an F-Score of 93.37%.This research also explores the challenges of hyperparameter optimization in deep learning models using swarm intelligence-based optimizers. These optimizers, inspired by natural behaviors, facilitate problem-solving through intelligent distribution, indirect interactions among agents, and simplification of complex processes. Additionally, the findings revealed that HBA outperformed GWO in determining optimal hyperparameter values, leading to enhanced model performance. Overall, the study highlights the potential of integrating deep learning models with swarm intelligence optimizers to improve sleep apnea detection.

ObjectiveThis study aims to present a novel model for predicting the future commitments of insurance companies that can adequately address the potential challenges of traditional methods. Traditionally, insurance companies use the Chain Ladder approach as a statistical tool to forecast the trend of claims development. This statistical method is favored by regulatory authorities in various countries due to its simplicity in assumptions and clear interpretation. However, certain assumptions, such as the stability of data development and linear relationships between variables, can affect the efficiency of this model when faced with internal policies or external factors like the COVID-19 pandemic. Forecasting future commitments close to reality is closely related to the financial stability of insurance companies. The amount that insurance companies allocate to meet their future obligations is identified as reserves. Calculating reserves that are less than the required amounts can pose challenges for insurance companies in fulfilling their commitments while calculating more than necessary amounts can negatively impact the financial statements of insurance companies. MethodsIn this study, a dynamic model based on machine learning algorithms is proposed. The model's output, which combines the number and timing of bodily injury accidents, plays a crucial role in calculating reserves for non-life insurance products. This model is specifically trained to predict the frequency of accidents in Vehicle Third-Party Liability Insurance. It can identify hidden patterns and non-linear, complex relationships within claims data. A Long Short-Term Memory (LSTM) neural network algorithm is employed, recognized for its strong predictive capability in time series data. The model is trained using historical data from Karafarin Insurance Company covering the years 2017 to 2021. ResultsThe performance of the model is highly related to the hyperparameters chosen for the model. Two of the most common approaches for tuning the hyperparameters are tested in this study. These Two models are grid and random search. The Root Mean Square Error (RMSE) is used as a performance metric, and it indicates that the grid search has a lower RMSE than the random search for the training data with a slight difference (16.33 versus 17.4). However, the results for the test data in the grid search have a sign of overfitting. ConclusionThis study recommends using random search for tuning the hyperparameters of the model to predict the frequency of daily incidents. The evaluation of the two approaches for tuning hyperparameters indicates that random search is more suitable for working with unfamiliar data and managing overfitting situations. Overfitting occurs when the model becomes overly influenced by the training data, learning not only the actual patterns but also the noise and minor details of the data. This issue can negatively impact the model's generalization ability.