Similar Tasks Research Articles

Academic procrastination, task values, and self-efficacy of pre-service biology teachers in online learning based on a gender perspective

Online learning can contribute to increased academic procrastination, often linked to low self-efficacy and perceived task value. This study investigates academic procrastination, self-efficacy, and task values from a gender perspective using a cross-sectional approach with students in their 1st, 3rd, 5th, and 7th semesters. An independent t-test assessed differences between male and female students. Findings showed no significant differences in academic procrastination across genders in any semester; however, notable differences in self-efficacy and task value emerged only in the 5th semester. The lack of significant differences in procrastination is attributed to similarities in environmental, psychological, and social factors, alongside common challenges in time management and distractions. Male and female students demonstrated comparable self-efficacy due to shared academic settings, social support, and media influences, while similar task values stemmed from common interests and experiences. The differences in the 5th semester likely arise from increased academic demands and evolving social dynamics, including gender stereotypes. Future research should employ qualitative methods, such as interviews or focus groups, to gain deeper insights into student experiences. Longitudinal studies could also track changes in self-efficacy and task value over time for a more comprehensive understanding. Targeted interventions to enhance self-efficacy and task value could be particularly beneficial in semesters with significant differences, thereby addressing academic procrastination more effectively.

What matters in teaching for students’ learning opportunities of subtraction in the 1–20 number range?

ABSTRACT Not all students in early grades develop efficient strategies for solving subtraction tasks. In this paper, we examine subtraction teaching in the 1–20 number range. We analyzed two first-grade lessons addressing similar subtraction tasks, using variation theory to identify what aspects of the content were foregrounded in the teaching. The analysis showed that both lessons supported the discernment of aspects of subtraction. However, the learning opportunities differed depending on what aspects the students were enabled to discern, the recurrence and variation of the elicited aspects, and how the teachers made the aspects visible by means of representations. The findings highlight how teaching can make it more likely for students to experience aspects of learning necessary to solve subtraction tasks such as 13–5 = _ . Additionally, the findings show how subtraction can be taught to enhance strategies using number relations that are also useful in higher number ranges.

Jack of all trades, master of one: domain-specific and domain-general contributions to perceptual expertise in visual comparison

Perceptual expertise is typically domain-specific and rarely generalises beyond an expert’s domain of experience. Forensic feature-comparison examiners outperform the norm in domain-specific visual comparison, but emerging research suggests that they show advantages on other similar tasks outside their domain of expertise. For example, fingerprint examiners not only outperform novices in fingerprint comparison, but also in face comparison. Yet, the extent to which their skills generalise is poorly understood. In this study, we investigated the generalisability of perceptual expertise amongst forensic examiners by comparing their performance to novices and other examiners within and outside their area of expertise. We recruited 85 experts from three forensic disciplines (face, fingerprint, and firearms) and asked them to complete four different visual comparison tasks: faces, fingerprints, firearms, and novel-objects. Examiners displayed domain-specific expertise: they outperformed novices and other examiners within their domain of visual comparison expertise. Yet, some of their skill also generalised: examiners also outperformed novices outside their area of expertise. However, while individual differences in examiners’ performance within their domain of experience were associated with their performance in a novel comparison task, they were not related to their performance on tasks outside their expert domain. These results provide key insight into the domain-specific and domain-general contributions of forensic examiners’ perceptual expertise. Forensic expertise lends some generalisable skill to other visual comparison tasks, but best performance is still seen within examiners’ domain of expertise.

Predicting food consumption using contextual factors: An application of machine learning models

Abstract Background Improving diet quality relies on making manageable adjustments to eating behaviours. Personalised nutrition interventions hold promise for modifying behaviour. Machine learning (ML) offers a novel approach to examining dietary behaviours in personalised nutrition by leveraging data on past behaviours and environmental contexts. This study aims to investigate whether contextual factors at eating occasions (EO) can predict food consumption to enhance diet quality. Methods Cross sectional data from the Measuring Eating in Everyday Life Study (MEALS) were analysed (n = 675, 18-35y). A smartphone food diary app recorded dietary intakes at EO for 3-4 non-consecutive days, also capturing social-environmental (e.g., activity) and physical-environmental factors (e.g., consumption location). Participant characteristics were collected via an online survey. Food groups intake (servings per EO) followed Australian Dietary Guidelines. This study benchmarked two established models, gradient boost decision tree and random forest, which have previously shown high performance in similar tasks. Performance was evaluated using 10-fold cross-validation, measuring mean absolute error (MEA), root mean square error (RMSE), and R squared. Feature importance analysis identified key variables for predicting food consumption. Results ML predicts most food groups at EO using contextual factors, with slight differences between actual and predicted consumption (<1 serving per EO). For fruits, dairy, and meat, MEA values were 0.35, 0.34, and 0.56 servings, respectively (RMSE values: 0.61, 0.50, and 0.80 servings). Self-efficacy, age and consumption location were influential in most ML models. Conclusions ML offers insights into contextual factors and food consumption, suggesting directions for precision nutrition interventions. Future research should identify positive influences of contextual factors on dietary behaviours and incorporate these insights into interventions. Key messages • Machine learning can effectively predict food consumption based on contextual factors at eating occasions. • Understanding the influence of contextual factors such as consumption location on food consumption can inform the development of precision nutrition interventions aimed at improving diet quality.

A Method for Teaching Biostatisticians to Write and Critique Specific Aims

This article describes the design of a single instructional module on writing and critiquing the Specific Aims (SA) section of a grant proposal for a biomedical research study. It begins by discussing the goals and learning objectives for the module. Then, it discusses the typical structure of a SA section to orient readers to the core content. This method is rooted in using an established framework for scientific writing, which is presented in brief, along with examples of how it applies to writing the SA section. Then, the structure of the instructional module is described, including the exercises for participants. Finally, the utility of this approach for teaching an otherwise difficult-to-teach competency is discussed, as well as how this approach might be extended to similar pedagogic tasks.

Meta-learning of human motor adaptation via the dorsal premotor cortex

Meta-learning enables us to learn how to learn the same or similar tasks more efficiently. Decision-making literature theorizes that a prefrontal network, including the orbitofrontal and anterior cingulate cortices, underlies meta-learning of decision making by reinforcement learning. Recently, computationally similar meta-learning has been theorized and empirically demonstrated in motor adaptation. However, it remains unclear whether meta-learning of motor adaptation also relies on a prefrontal network. Considering hierarchical information flow from the prefrontal to motor cortices, this study explores whether meta-learning is processed in the dorsolateral prefrontal cortex (DLPFC) or in the dorsal premotor cortex (PMd), which is situated upstream of the primary motor cortex, but downstream of the DLPFC. Transcranial magnetic stimulation (TMS) was delivered to either PMd or DLPFC during a motor meta-learning task, in which human participants were trained to regulate the rate and retention of motor adaptation to maximize rewards. While motor adaptation itself was intact, TMS to PMd, but not DLPFC, attenuated meta-learning, impairing the ability to regulate motor adaptation to maximize rewards. Further analyses revealed that TMS to PMd attenuated meta-learning of memory retention. These results suggest that meta-learning of motor adaptation relies more on the premotor area than on a prefrontal network. Thus, while PMd is traditionally viewed as crucial for planning motor actions, this study suggests that PMd is also crucial for meta-learning of motor adaptation, processing goal-directed planning of how long motor memory should be retained to fit the long-term goal of motor adaptation.

Discrimination of mnemonic similarity is associated with short-term and long-term memory precision.

Remembering specific memories with precision relies on the differentiation of similar memory contents - a process commonly referred to as pattern separation and behaviorally operationalized as lure discrimination in the mnemonic similarity task. Although pattern separation is typically investigated in the context of long-term memory (LTM), recent research extends these findings to short-term memory (STM) within a mixture model framework, emphasizing the distinction between memory quality and quantity. According to this framework, pattern separation is associated with memory precision across STM and LTM, regardless of the overall memory likelihood. However, these associations among memory quality measures may persist without the mixture model assumption. In an alternative model, a unitary memory strength measure quantified as a discrimination score (d') may also capture the association between pattern separation and memory quality, as pattern separation has been previously linked with strength-based memory performance. We tested these possibilities based on individual differences among 132 participants who underwent tasks measuring mnemonic pattern separation and STM/LTM quality. We found that behavioral estimates of pattern separation are significantly correlated with STM and LTM precision, irrespective of the likelihood of STM/LTM recall success. However, these associations are absent when considering the correlation between pattern separation and memory strength under a unitary model framework. By leveraging individual differences to constrain our understanding of cognitive models, our data unravel the intricate relationship between pattern separation and memory quality across timescales. These findings may therefore contribute to identifying sensitive behavioral measures for detecting subtle memory deficits in older adults or clinical populations.

The Primacy of Commercial Gain over Ideology in Wars and Armed Conflicts

The paper provides a brief analysis of the causes and reasons for military conflicts and wars in the past and present and makes a forecast for the future. By comparing ideological differences and commercial goals, the author concludes that economic priorities are the primary causes of military conflicts and wars. Analyzing historical examples, the author shows that economic priorities have always been the main reasons for unleashing military actions. The main objectives that the aggressor solved by unleashing war were: expansion of the territory of its habitat; plundering of the defeated; transformation of the defeated population into slaves (cheap labor). These priorities, in the author’s opinion, are the main ones in all wars and armed conflicts in history. Similar tasks are currently being solved by the industrial and banking oligarchs of the United States and the United West in the Middle East, Africa, Latin America and Europe. Other goals and objectives of wars and armed conflicts are secondary and can be realized as additional ones. In this regard, the question of the primary and secondary objectives of commercial benefit and ideology is resolved in favor of commercial benefit. In the paper, on the basis of scientific and research works of the author, it is concluded that the current situation in the world is a well thought out and carefully calculated by the financial oligarchs of the United States and the united West, when a commercial project and its realization should bring them substantial profits, and the population of the planet Earth — destruction, grief and suffering.

Semantic Alignment in Disciplinary Tutoring System: Leveraging Sentence Transformer Technology

In this work, we present a disciplinary e-tutoring system that integrates ONTO-TDM, an ontology designed for teaching domain modeling, with advanced transformer technology. Our primary objective is to enhance semantic similarity tasks within the system by fine-tuning a Sentence Transformer model. By carefully adjusting training parameters with a curated dataset of question-answer pairs focused on algorithms and data structures, we achieved a notable improvement in system performance. The Sentence Transformer model, combined with domain ontology, achieved an accuracy of 91%, a precision of 93%, a recall of 89%, and an F1-score of 90%, significantly surpassing the results of existing works. This methodology highlights the potential to deliver personalized support and guidance in tutoring scenarios. It effectively addresses the evolving needs of modern education by offering tailored answers and reducing the necessity for constant learner-tutor interaction, thereby improving the efficiency of educational support systems.

Maternal trauma exposure is linked to adult offspring's cognitive functioning in the context of mass violence.

This study aimed to examine whether mothers' level of trauma exposure and posttraumatic stress disorder (PTSD) symptoms are related to their offspring's cognitive functioning. Mothers exposed to the 1994 genocide perpetrated against the Tutsi (N = 181) and one of their adult offspring were recruited in Rwanda. Mothers and their offspring answered questionnaires on sociodemographic information, the level of trauma exposure, and PTSD symptoms. They also performed a backward digit span task and a similarity task to assess their working memory and abstract reasoning, respectively. Hierarchical regression models were used to analyze the data. Mothers' level of trauma exposure, but not PTSD symptoms, was significantly related to their offspring's abstract reasoning performance after controlling for offspring's age, sex, and economic status. The relationship between mothers' level of trauma exposure, PTSD symptoms, and their offspring's working memory performance was not significant. Offspring's education was significantly linked to their working memory and abstract reasoning performance. The findings demonstrate that the intergenerational transmission of trauma can impact the cognitive functioning of the next generation. These results also illustrate the profound and long-term human impact of mass violence. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Oxyanion Engineering on RuO2 for Efficient Proton Exchange Membrane Water Electrolysis.

In acidic proton exchange membrane water electrolysis (PEMWE), the anode oxygen evolution reaction (OER) catalysts rely heavily on the expensive and scarce iridium-based materials. Ruthenium dioxide (RuO2) with lower price and higher OER activity, has been explored for the similar task, but has been restricted by the poor stability. Herein, we developed an anion modification strategy to improve the OER performance of RuO2 in acidic media. The designed multicomponent catalyst based on sulfate anchored on RuO2/MoO3 displays a low overpotential of 190 mV at 10 mA cm-2 and stably operates for 500 hours with a very low degradation rate of 20 μV h-1 in acidic electrolyte. When assembled in a PEMWE cell, this catalyst as an anode shows an excellent stability at 500 mA cm-2 for 150 h. Experimental and theoretical results revealed that MoO3 could stabilize sulfate anion on RuO2 surface to suppress its leaching during OER. Such MoO3-anchored sulfate not only reduces the formation energy of *OOH intermediate on RuO2, but also impedes both the surface Ru and lattice oxygen loss, thereby achieving the high OER activity and exceptional durability.

Effective Hard Negative Mining for Contrastive Learning-based Code Search

Background. Code search aims to find the most relevant code snippet in a large codebase based on a given natural language query. An accurate code search engine can increase code reuse and improve programming efficiency. The focus of code search is how to represent the semantic similarity of code and query. With the development of code pre-trained models, the pattern of using numeric feature vectors (embeddings) to represent code semantics and using vector distance to represent semantic similarity has replaced traditional string matching methods. The quality of semantic representations is critical to the effectiveness of downstream tasks such as code search. Currently, the state-of-the-art (SOTA) learning method uses the contrastive learning paradigm. The objective of contrastive learning is to maximize the similarity between matching code and query (positive samples) and minimize the similarity between mismatched pairs (negative samples). To increase the reusing of negative samples, prior contrastive learning approaches use a large queue (memory bank) to store embeddings. Problem. However, there is still a lot of room for improvement in using negative examples for code search: ① Due to the random selection of negative samples, semantic representations learned by existing models cannot distinguish similar codes well. ② Since semantic vectors in the memory bank are reused from previous inference results and then directly used for loss function calculation without gradient descent, the model cannot effectively learn the negative sample semantic information. Method. To solve the above problems, we propose a contrastive learning code search model with hard negative mining called CoCoHaNeRe: ❶ To enable the model to distinguish similar codes, we introduce hard negative examples into contrastive training, which are negative examples in the codebase that are most similar to positive examples. As a result, hard negative examples are most likely to make the model make mistakes. ❷ To improve the learning efficiency of negative samples during training, we add all hard negative examples to the model's gradient descent process. Result. To verify the effectiveness of CoCoHaNeRe, we conducted experiments on large code search datasets with six programming languages, as well as similar retrieval tasks code clone detection and code question answering. Experimental results show that our model achieves SOTA performance. In the code search task, the average MRR score of CoCoHaNeRe exceeds CodeBERT, GraphCodeBERT, and UniXcoder by 11.25%, 8.13%, and 7.38%, respectively. It has also made great progress in code clone detection and code question answering. In addition, our method performs well in different programming languages and code pre-training models. Furthermore, qualitative analysis shows that our model effectively distinguishes high-order semantic differences between similar codes.

Soundscape and indoor air quality in naturally ventilated educational environments: a multi-domain study

In this study, indoor air quality (IAQ) and acoustics were jointly analysed to determine their i) combined effect on students' performance in a math task, and ii) cross-modal effect on soundscape perception. 192 students (12 to 14 years) took part in a laboratory experiment. Students were separated into two classroom-like chambers with different IAQ conditions: a CO2 concentration of 800 and 2000 ppm, respectively. The participants were asked to solve calculations and answer a soundscape questionnaire in three listening conditions: quiet, a natural scenario with birdsongs, and one scenario with children's playground activity sounds. In the latter two conditions, the sounds were played back through headphones at a level of 47 dBA. Results show that students had faster response times in the math task in quiet compared to the listening condition with children's activity sounds, while maintaining a similar task accuracy. No effect of the IAQ conditions and natural sounds was found on the task performance. No cross-modal effect of CO2 concentration on soundscape perception was detected. Regarding soundscape assessments, playground activity was rated as the most annoying listening condition, while listening to birdsongs was the most pleasant one, despite its noise level exceeding the quiet condition.

A study of noise-induced annoyance and fatigue in open-plan offices: comparison between hearing-impaired and normal-hearing listeners

Several studies outline a future scenario of longer working lives and later retirement, meaning that there could be a large number of elderly workers in offices. Add to this the natural effect of hearing loss due to ageing (presbycusis; which usually presents after the 5th decade of life) and you have an open door to questions regarding the acoustic comfort of this group. In a first study, it was shown that the effect of intelligible speech on performance in a serial recall task is similar for listeners with normal hearing and presbycusis. The aim of this new experiment is to assess whether any differences might occur when people are involved in a more ecologically valid task (long exposure to noise - 3 h/condition - and similar tasks to those performed in open-plan offices). Given the validation of a hearing loss simulator in previous studies, this one will be used to reproduce the group of people with mild hearing loss (onset of presbycusis). Therefore, the same group of people (32 participants) with normal hearing will perform the experiment under two listening conditions, with and without the simulator. Each listening condition will have two noise conditions, irrelevant speech, each with a different intelligibility level.

Enhancing Social-Communication and Behavior Skills in Adolescents with Intellectual Disabilities through Computerized Cognitive Training

Although it is not disputed that cognitive training can affect similar domains (near transfer) such as similar working memory tasks, it is still hotly debated whether they advance seemingly different domains (far transfer), such as social skills improvement. This study aimed to investigate the possibility of far transfer in enhancing social-communication and adaptive behavior skills using a computerized cognitive training program (CCTP) in adolescents with intellectual disabilities (ID). The two research participants aged 14 & 16 conducted a 16-week home-based CCTP using CogniFit for 2-4 days a week, at 20 minutes a day. After the 16-week CCTP, an adjusted 4-week CCTP based on CogniFit activities and on added social cues was followed. The participants underwent pre and post assessments for examining far transfer skills using a) the Test of Pragmatic Language-2, b) the Social Skills Improvement System Rating Scale, and c) the Vineland Adaptive Behavior Scales II. The results showed that this intervention program can benefit adolescents with ID in social-communication and adaptive skills, which can function as an important driver of cognitive development and academic outcomes.

A depth graph attention-based multi-channel transfer learning network for fluid classification from logging data

Fluid classification is a fundamental task in the field of geological sciences to achieve effective reservoir characterization and hydrocarbon exploration. Traditional fluid classification methods are often limited by long processing times and an inability to capture complex relationships within the data. To address this issue, this paper proposes a novel deep learning approach—the Deep Graph Attention Multi-channel Transfer Learning Network (DGMT), aimed at improving the efficiency and accuracy of fluid classification from logging data. This model comprises three key components: a graph attention layer, a multi-channel feature extractor, and a transfer learning module. The graph attention layer is designed to handle spatial dependencies between different logging channels, enhancing classification accuracy by focusing on critical features. The multi-channel feature extractor integrates information from various data sources, ensuring comprehensive utilization of the rich information in logging data. The transfer learning module allows the model to transfer knowledge from pre-trained models of similar tasks, accelerating the training process and significantly improving the model's generalization ability and robustness. This feature enables the DGMT model to adapt to different geological environments and logging conditions, showing superior performance over traditional methods. To validate the effectiveness of the DGMT model, we conducted experiments on actual logging datasets containing multiple oil wells. The experimental results indicate that, compared to common machine learning algorithms and other deep learning methods, the DGMT model significantly improves in accuracy and other classification performance metrics.

Enhancing Cloud Task Scheduling with Multi-Objective Optimization Using K-Means Clustering and Dynamic Resource Allocation

The scheduling and resource allocation procedure is an essential component of cloud resource management. Effective resource allocation is severely hampered by the task arrival rates' erratic and unclear behavior. To prevent under or overusing resources, an effective scheduling strategy is necessary. To improve scheduling and allocation performance, a multi-objective optimization technique is presented for the best resource allocation and task scheduling inside scientific workflow datasets in a heterogeneous environment. In the first stage, the system calculates four key metrics: Communication Cost, Computation Cost, Earliest Finished Time on a particular VM, and Total Task Length for a specific scientific workflow dataset. These metrics provide a comprehensive understanding of the resource requirements and help make informed scheduling decisions. In the second stage, tasks are clustered using the K-Means clustering algorithm. This clustering groups similar tasks together, making managing and scheduling them easier. In the third stage, the proposed resource allocation algorithm allocates the clustered tasks to the appropriate VMs. This step ensures that the tasks are assigned to the best-suited resources, optimizing the overall system performance and resource utilization. By following this multi-stage process, the system aims to achieve optimal resource allocation and task scheduling, thereby improving the efficiency and effectiveness of cloud resource management. The proposed method significantly outperforms PSO, CSO, and GWO by consistently achieving lower Makespan—under 400 units at 50 tasks—while maintaining high resource utilization rates above 0.75, demonstrating superior efficiency in task execution and resource management.

Increasing Generative Artificial Intelligence Competency among Students Enrolled in Doctoral Nursing Research Coursework.

Generative artificial intelligence (AI) tools may soon be integrated into health care practice and research. Nurses in leadership roles, many of whom are doctorally prepared, will need to determine whether and how to integrate them in a safe and useful way. This study aimed to develop and evaluate a brief intervention to increase PhD nursing students' knowledge of appropriate applications for using generative AI tools in health care. We created didactic lectures and laboratory-based activities to introduce generative AI to students enrolled in a nursing PhD data science and visualization course. Students were provided with a subscription to Chat Generative Pretrained Transformer (ChatGPT) 4.0, a general-purpose generative AI tool, for use in and outside the class. During the didactic portion, we described generative AI and its current and potential future applications in health care, including examples of appropriate and inappropriate applications. In the laboratory sessions, students were given three tasks representing different use cases of generative AI in health care practice and research (clinical decision support, patient decision support, and scientific communication) and asked to engage with ChatGPT on each. Students (n = 10) independently wrote a brief reflection for each task evaluating safety (accuracy, hallucinations) and usability (ease of use, usefulness, and intention to use in the future). Reflections were analyzed using directed content analysis. Students were able to identify the strengths and limitations of ChatGPT in completing all three tasks and developed opinions on whether they would feel comfortable using ChatGPT for similar tasks in the future. All of them reported increasing their self-rated competency in generative AI by one to two points on a five-point rating scale. This brief educational intervention supported doctoral nursing students in understanding the appropriate uses of ChatGPT, which may support their ability to appraise and use these tools in their future work.

Context-Based Meta-Reinforcement Learning With Bayesian Nonparametric Models.

Deep reinforcement learning agents usually need to collect a large number of interactions to solve a single task. In contrast, meta-reinforcement learning (meta-RL) aims to quickly adapt to new tasks using a small amount of experience by leveraging the knowledge from training on a set of similar tasks. State-of-the-art context-based meta-RL algorithms use the context to encode the task information and train a policy conditioned on the inferred latent task encoding. However, most recent works are limited to parametric tasks, where a handful of variables control the full variation in the task distribution, and also failed to work in non-stationary environments due to the few-shot adaptation setting. To address those limitations, we propose MEta-reinforcement Learning with Task Self-discovery (MELTS), which adaptively learns qualitatively different nonparametric tasks and adapts to new tasks in a zero-shot manner. We introduce a novel deep clustering framework (DPMM-VAE) based on an infinite mixture of Gaussians, which combines the Dirichlet process mixture model (DPMM) and the variational autoencoder (VAE), to simultaneously learn task representations and cluster the tasks in a self-adaptive way. Integrating DPMM-VAE into MELTS enables it to adaptively discover the multi-modal structure of the nonparametric task distribution, which previous methods using isotropic Gaussian random variables cannot model. In addition, we propose a zero-shot adaptation mechanism and a recurrence-based context encoding strategy to improve the data efficiency and make our algorithm applicable in non-stationary environments. On various continuous control tasks with both parametric and nonparametric variations, our algorithm produces a more structured and self-adaptive task latent space and also achieves superior sample efficiency and asymptotic performance compared with state-of-the-art meta-RL algorithms.

FORECASTING AND OPTIMIZATION OF CATALYTIC CRACKING UNIT OPERATION UNDER CONDITIONS OF FUZZY INFORMATION

This paper discusses the application of nonlinear regression to forecast and optimize the operation of catalytic cracking units under conditions of fuzzy information. Catalytic cracking is a crucial process in oil refining that produces high-quality gasoline and other light hydrocarbon products. However, the complexity of the process and the uncertainty of initial data complicate the modeling and optimization of plant operations. To address this issue, a nonlinear regression method is proposed that accommodates the fuzziness of input and output parameters described by linguistic variables. The methodology includes the collection and formalization of expert knowledge, the construction of fuzzy models, and their integration into the process control system. Forecasting is performed by creating regression models that describe the relationships between operational parameters and product quality characteristics. The paper presents a procedure for developing and applying nonlinear regression models, describes algorithms for synthesizing linguistic models, and provides examples of their use to optimize the operation of catalytic cracking units. The modeling results demonstrate the high adequacy and accuracy of the proposed method, as well as its advantages over traditional approaches in conditions of uncertainty and data scarcity. The scientific novelty of the research lies in the development and testing of advanced nonlinear regression models adapted for analyzing and optimizing catalytic cracking processes based on fuzzy data. These methods take into account the specificity and uncertainty of process data, improving the accuracy and reliability of forecasts, which facilitates more effective management of production processes in the petrochemical industry. The main reason for conducting this study is the need to improve the control of oil refining processes, particularly catalytic cracking, which plays an important role in producing high-quality gasoline. The complexity of this process and the presence of fuzzy information caused by fuzzy initial data require the development of new modeling and optimization methods. Existing traditional models based on deterministic methods are often insufficient under uncertainty. This leads to a decrease in the accuracy of process control, which can negatively affect the quality of the final product and production efficiency. The use of nonlinear regression in combination with fuzzy logic is a more flexible and adaptive approach that allows you to take into account the fuzziness and uncertainty of data and use expert knowledge to build models that match the actual operating conditions of the units. Thus, this study aims to solve the key problems associated with data uncertainty and the complexity of the catalytic cracking process, which will improve the accuracy of forecasting and optimization of the units. The main contribution is creating a model that uses nonlinear regression methods in combination with fuzzy logic. This allows uncertainty in input data (such as reactor temperature or pressure) to be effectively considered and processed to improve gasoline and other product yield forecasts. It is shown that using nonlinear regression combined with fuzzy logic significantly improves the management of technological processes, increases the output and quality of products, and reduces production costs. The conclusion of the paper discusses the prospects for further development of the methodology and its application to solve similar tasks in other areas of chemical technology.