Convolutional Neural Networks (CNNs) are a kind of artificial neural network designed to extract features and find out patterns for tasks such as segmentation, recognizing objects, and drawing up classification. Within a CNNs architecture, pooling operations are used until the number of parameters and the computational complexity are reduced. Numerous papers have focused on investigating the impact of pooling on the performance of Convolutional Neural Networks (CNNs), leading to the development of various pooling models. Recently, a fuzzy pooling operation based on type-1 fuzzy sets was introduced to cope with the local imprecision of the feature maps. However, in fuzzy set theory, it is not always accurate to assume that the degree of non-membership of an element in a fuzzy set is simply the complement of the degree of membership. This is due to the potential existence of a hesitation degree, which implies a certain level of uncertainty. To overcome this limitation, intuitionistic fuzzy sets (IFS) were introduced to incorporate the concept of a degree of hesitation. In this paper, we introduce a novel pooling operation based on intuitionistic fuzzy sets to incorporate the degree of hesitation heretofore neglected by a fuzzy pooling operation based on classical fuzzy sets, and we investigate its performance in the context of image classification. Intuitionistic pooling is performed in four steps: bifuzzification (by the transformation of data through the use of membership and non-membership maps), first aggregation (through the transformation of the IFS into a standard fuzzy set, second aggregation (through the transformation and use of a sum operator), and the defuzzification of feature map neighborhoods by using a max operator. IFS pooling is used for the construction of an intuitionistic pooling layer that can be applied as a drop-in replacement for the current, fuzzy (type-1) and crisp, pooling layers of CNN architectures. Various experiments involving multiple datasets demonstrate that an IFS-based pooling can enhance the classification performance of a CNN. A benchmarking study reveals that this significantly outperforms even the most recent pooling models, especially in stochastic environments.
6 January 2025 in Revista de Gestão Social e Ambiental
Objective: The objective of this study is to develop an efficient glucose management system based on an original control model with mixed constraints (dynamic and static constraints) to manage diets with several thresholds, using Bergman’s minimal model and Fischer’s meal models. Theoretical Framework: Personalized and healthy diets play a significant role in controlling the progression of diseases such as type 2 diabetes. Existing glucose management methods are predominantly based on dynamic mathemati- cal models and focus on diets with a single peak. However, diets with multiple thresholds better reflect realistic dietary patterns and enable more comprehen- sive glucose control. Method: The proposed model combines two key components: an objective function that minimizes the total error between the basic and generated glucose and the quantity of diet, and constraints that incorporate dynamic differential equations (Bergman’s minimal model) and static constraints representing the interval between two sensitive meals. To solve this model, the integral in the objective function is discretized using the trapezoidal rule, and dynamic con- straints are approximated using the Euler-Cauchy method. A genetic algorithm is employed to optimize the resulting static model. Several study cases, including global, partial, and one-off scenarios, are considered. Results and Discussion: Numerical simulations demonstrate that the pro- posed system effectively estimates sufficient personalized diets, grading the glycemic profile within a defined confidence zone. This approach accounts for three critical dimensions: the quantity of food, its distribution, and the intervals between successive meals. Optimal management of these parameters results in maximum dietary intake, low glucose generation, and minimal insulin secretion, enabling diabetics to maintain a near-normal lifestyle. Research Implications: This study provides a realistic and innovative frame- work for glucose management, considering the complexities of real-life dietary patterns. It highlights the significance of integrating dynamic and static con- straints in glucose control systems and contributes to the design of personalized dietary strategies for individuals with diabetes. Originality/Value: This work contributes to the literature by introducing a novel glucose management system that considers diets with multiple thresholds, enabling better control of glycemic profiles and improving the quality of life for individuals with diabetes.