Fuzzy clustering methods are conveniently employed in constructing a fuzzy model of a system, but they need to tune some parameters. In this research, FCM is chosen for fuzzy clustering. Parameters such as the number of clusters and the value of fuzzifier significantly influence the extent of generalization of the fuzzy model. These two parameters require tuning to reduce the Overfitting in the fuzzy model. Two new cost functions are developed to set the parameters of FCM algorithm properly and the two evolutionary optimization algorithms, i.e. the multi-objective simulated annealing and the multi-objective imperialist competitive algorithm, are employed to optimize the parameters of FCM according to the proposed cost functions. The multi-objective imperialist competitive algorithm is the proposed algorithm.

In predictive modeling, Overfitting poses a significant risk, particularly when the feature count surpasses the number of observations, a common scenario in highdimensional datasets. To mitigate this risk, feature selection is employed to enhance model generalizability by reducing the dimensionality of the data. This study evaluates the stability of feature selection techniques with respect to varying data volumes, focusing on time series similarity methods. Utilizing a comprehensive dataset that includes the closing, opening, high, and low prices of stocks from 100 high-income companies listed in the Fortune Global 500, this research compares several feature selection methods, including variance thresholds, edit distance, and Hausdorff distance metrics. Numerous feature selection methods were investigated in literature. Selecting the more accurate feature selection methods in order to forecast can be challenging [1]. So, this study examines the most well-known feature selection methods’ performance in different data sizes. The aim is to identify methods that show minimal sensitivity to the quantity of data, ensuring robustness and reliability in predictions, which is crucial for financial forecasting. Results indicate that among the tested feature selection strategies, the variance method, edit distance, and Hausdorff methods exhibit the least sensitivity to changes in data volume. These methods, therefore, provide a dependable approach to reducing feature space without significantly compromising predictive accuracy. This study highlights the effectiveness of time series similarity methods in feature selection and underlines their potential in applications involving fluctuating datasets, such as financial markets or dynamic economic conditions.

In order to study the effect of R2O/Al2O3 (where R=Na or K), SiO2/Al2O3, Na2O/K2O and H2O/R2O molar ratios on the compressive strength (CS) of Metakaolin base geopolymers, more than forty data were gathered from literature. To increase the number of data, some experiments were also designed. The resulted data were utilized to train and test the three layer artificial neural network (ANN). Bayesian regularization method and Early Stopping methods with back propagation algorithm were applied as training algorithm. Good validation for CS was resulted due to the inhibition of Overfitting problems with the applied training algorithm. The results showed that optimized condition of SiO2/Al2O3, R2O/Al2O3, Na2O/K2O and H2O/R2O ratios to achieve high CS should be 3.6-3.8, 1.0-1.2, 0.6-1 and 10-11, respectively. These results are in agreement with probable mechanism of geopolymerization.

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.

In this research, a method based on the optimal neural network and pattern search optimization algorithm is presented to solve reliability-based design optimization problems. The main idea is to find a surrogate model that does not suffer from the phenomenon of Overfitting and therefore has a good generalization accuracy. In the first stage, using the program written using the SM toolbox, a data set of inputs and outputs of the problem is created by running Sap2000. Then an optimization problem is solved to obtain the best performance of the neural network. The design variables in the neural network training stage, are the number of layers, the number of neurons in each layer and the type of transfer. The objective function is the performance ratio, which is defined as the ratio of the number of parameters in the neural network to the number of members of the data set used in the training process. Subsequently. The pattern search algorithm is used to optimize the examples using the developed optimal ANN as a surrogate model. To demonstrate the effectiveness of the presented method, two numerical examples are considered. In the first example, a ten-bar plane truss and in the second example, a two-layered 832-membered barrel vault have been investigated. In the first example, the proposed method has worked about 32 times faster in the vase of continuous variables and 25 times faster in the case of discrete variables. (Compared to solving the problem with the original SAP 2000 model and the Latin hypercube sampling method). In both examples, the surrogate model obtained from the proposed method has provided the desired performance in both the validation and the test data.

The segmentation of retinal vessels is vital for timely diagnosis. and treatment of various eye diseases. However, due to inherent characteristics of retinal vessels in fundus images such as changes in thickness, direction, and complexity of vessels, as well as imbalanced contrast between background and vessels, segmenting retinal vessels continues to pose significant challenges. Also, despite advancements in CNN-based methods, challenges such as insufficient extraction of structural information, complexity, Overfitting, preference for local information, and poor performance in noisy conditions persist. To address these drawbacks, in this paper we proposed a novel modified U-Net named DABT-U-Net. Our method enhances discriminative capability by introducing Hierarchical Dilated Convolution (HDC), Dual Attentive BConvLSTM, and Multi-Head Self-Attention (MHSA) blocks. Additionally, we adopt a collaborative patch-based training approach to mitigate data scarcity and Overfitting. Evaluation on the DRIVE and STARE datasets shows that DABT-U-Net achieves superior accuracy, sensitivity, and F1 score compared to existing methods, demonstrating its effectiveness in retinal vessel segmentation. Specifically, our proposed method demonstrates improvements in accuracy, sensitivity, and F1 score by 0.32%, 0.61%, and 0.14%, respectively, on the DRIVE dataset, and by 0.07%, 0.83%, and 0.14% on the STARE dataset compared to a less effective approach.

In recent years, deep learning methods based on the convolutional neural network (CNNs) have demonstrated good performance for hyperspectral image classification (HSI). Although, in order to obtain good results, we need a large number of training data in the CNNs to avoid the Overfitting problem. This paper aims to establish a segmentation-based method to extend the training data for deep learning-based hyperspectral image classification. First, two unsupervised segmentation methods (K-Means and Multi-resolution) are used for the segmentation of the hyperspectral images. Second, we obtained pseudo-training data which depends on the overlay between segmented hyperspectral images and original training data sets. So we extend the number of training samples for CNN to avoid the Overfitting problem and achieve good results. Finally, a Hybrid-CNN model that is a combination of 2D-Convolution and 3D-Convolution is applied to classify hyperspectral datasets with the training samples consisting of the original and pseudo training sets. The proposed method was tested on two Kennedy Space Center (KSC) and Botswana hyperspectral images and the results are compared with the two methods. The overall accuracy with the proposed method retrieves 100% and 96.11% for KSC and Botswana datasets, respectively. Also, we tested the proposed Hybrid-CNN network with Pavia University data, and the classification results show that the proposed Hybrid-CNN has good performance in the face of complex data. The overall accuracy retrieves 99.66% for the Pavia dataset. Keywords: Hyperspectral Image Classification (HSI), Convolutional Neural Network (CNN), Multiresolution segmentation, K-Means Clustering.

Background: Observation, categorize and count various types of white blood cells in a blood sample is a One of the most important steps in the treatment of various diseases. The aim of this study was to design and implement a fast and reliable and based on the processing of microscopic images of blood samples for the classification of four types of white blood cells. Materials and Methods: In this article, the modified k-means clustering method is used to perform image segmentation. Furthermore, The classification of white blood cells was done using a deep convolutional neural network and with the help of data in the MISP database, a free database composed of microscopic blood sample images. Moreover, Several regularization techniques such as dropout and image augmentation were applied to prevent the network from Overfitting. Results: In the classification category, the accuracy of the neural network is measured to be 99%, which has been more successful than many earlier studies. In the segmentation section, the cross-reference index was 0. 73. Conclusion: The results of this research show that rapid and reliable system design and implementation is possible by processing the microscopic images of the blood sample using different methods of image processing and machine learning.