Conventional Learning Models Research Articles

Pengaruh Penerapan Model Pembelajaran RBL Terhadap Kreativitas Belajar Siswa Kelas XI IPS Mata Pelajaran Ekonomi Di SMAN 2 Muaro Jambi

This research aims to determine the effect of applying conventional learning methods and the RBL learning model in increasing the learning creativity of class XI IPS students in economics subjects at SMAN 2 Muaro Jambi. The research design that will be carried out in this research is a quasi-experimental research with a pre-test – post-test control group design model with one type of treatment. The results of the research show that there are differences in the influence of the conventional learning model and the RBL learning model on the creativity of class XI IPS students. Students who learn using the RBL learning model obtain the best creative learning abilities.

Penerapan Model Pembelajaran Means-Ends Analysis Dalam Meningkatkan Pemahaman Siswa Pada Proses Pembelajaran Ekonomi Di Kelas XI SMAN 15 Muaro Jambi

This research aims to determine the application of conventional learning models in improving class XI students' understanding of economics subjects at SMAN 15 Muaro Jambi. The research design that will be carried out in this research is quasi-experimental research with a quantitative approach. The research results show that the application of the Means-Ends Analysis (MEA) learning model has proven effective in improving student learning outcomes in many aspects such as knowledge and skills. This is because MEA is hands-on and builds enthusiasm for learning so that the learning process can be student-centered rather than the role of the teacher

Production Prediction and Influencing Factors Analysis of Horizontal Well Plunger Gas Lift Based on Interpretable Machine Learning

With the development of unconventional natural gas resources, plunger gas lift technology has gained widespread application. Accurately predicting gas production from unconventional gas reservoirs is a crucial step in evaluating the effectiveness of plunger gas lift technology and optimizing its design. However, most existing prediction methods are mechanism-driven, incorporating numerous assumptions and simplifications that make it challenging to fully capture the complex physical processes involved in plunger gas lift technology, ultimately leading to significant errors in capacity prediction. Furthermore, engineering design factors and production system factors associated with plunger gas lift technology can contribute to substantial deviations in gas production forecasts. This study employs three powerful regression algorithms, XGBoost, Random Forest, and SVR, to predict gas production in plunger gas lift wells. This method comprehensively leverages various types of data, including collected engineering design, production system, and production data, directly extracting the underlying patterns within the data through machine learning algorithms to establish a prediction model for gas production in plunger gas lift wells. Among these, the XGBoost algorithm stands out due to its robustness and numerous advantages, such as high accuracy, ability to effectively handle outliers, and reduced risk of overfitting. The results indicate that the XGBoost algorithm exhibits impressive performance, achieving an R2 (coefficient of determination) value of 0.87 for six-fold cross-validation and 0.85 for the test set. Furthermore, to address the “black box” problem (the inability to know the internal working structure and workings of the model and to directly understand the decision-making process), which is commonly associated with conventional machine learning models, the SHAP (Shapley additive explanations) method was utilized to globally and locally interpret the established machine learning model, analyze the main factors (such as starting time of wells, gas–liquid ratio, catcher well inclination angle, etc.) influencing gas production, and enhance the credibility and transparency of the model. Taking plunger gas lift wells in southwest China as an example, the effectiveness and practicality of this method are demonstrated, providing reliable data support for shale gas production prediction, and offering valuable guidance for actual on-site production.

Job description parsing with explainable transformer based ensemble models to extract the technical and non-technical skills

The rapid digitization of the economy is transforming the job market, creating new roles and reshaping existing ones. As skill requirements evolve, identifying essential competencies becomes increasingly critical. This paper introduces a novel ensemble model that combines traditional and transformer-based neural networks to extract both technical and non-technical skills from job descriptions. A substantial dataset of job descriptions from reputable platforms was meticulously annotated for 22 IT roles. The model demonstrated superior performance in extracting both non-technical (67% F-score) and technical skills (72% F-score) compared to conventional CRF and hybrid deep learning models. Specifically, the proposed model outperformed these baselines by an average margin of 10% and 6%, respectively, for non-technical skills, and 29% and 6.8% for technical skills. A 5 × 2cv paired t-test confirmed the statistical significance of these improvements. In addition, to enhance model interpretability, Local Interpretable Model-Agnostic Explanations (LIME) were employed in the experiments.

Risk Assessment and Diagnosis Code Prediction from Electronic Health Records (EHR) using deep learning

Knowledge in the medical field is often expressed by distinct and subjective norms. Much research and development efforts have focused on using deep learning algorithms to predict the likelihood of illnesses from EHR in recent years. When it comes to risk prediction, deep learning-based techniques outperform more conventional machine learning models. But nothing in the literature fully accounts for what doctors already know, such as the connections between illnesses and their risk factors. This research examines the use of Multi-layer Perceptron models for the categorisation of diagnoses in electronic health records. The raw data and a modified version of the EHR dataset are used to train two MLPs with distinct topologies. For comparative purposes, a Random Forest is used as a baseline. To phenotype patients using their electronic health records, we provide a deep learning method. Predictive modelling of chronic illnesses is the particular scenario used to verify the suggested model on a real-world EHR data warehouse. Many deep learning applications on EHRs have been effective, and there is still a lot of potential to be realised. It has been discovered that deep learning models can learn from the limited EHR dataset, but not to a level where they outperform the baseline model.

Influence of Cognitive Style on Students' Critical Thinking Skills in Physics Material

This research aims to determine the influence of cognitive style on students' critical thinking abilities. This research uses a quasi-experimental method with a factorial design research design. The research subjects consisted of class XI MAN 1 Central Lombok students who were divided into experimental classes and control classes. The experimental class uses the PBL model, while the control class uses a conventional learning model. Research instruments include critical thinking ability tests and the Group Embedded Figures Test (GEFT) cognitive style test. Before the critical thinking ability test instrument is given, a test of the instrument is first carried out, namely a validity and reliability test. level of difficulty, and distinguishing power. The research results show that the significant level value is 0.001 which is smaller than α which is equal to 0.05 (0.000<0.05). These findings indicate that there is an influence of cognitive style on students' critical thinking abilities.

The effect of augmented reality-based breaststroke swimming learning model on deaf students

Education is an important need for humans, through education a person gains knowledge and abilities that can be used as provisions for a better life in the future. The research aims to determine the effectiveness of the Augmented Reality-based breaststroke swimming learning model for students with hearing impairments at State SLB in Solo Raya in 2023. This study was an experimental study. The research design used was non-equivalent control group design. The sample in this study amounted to 22 students. Data collection in this study used a breaststroke swimming skill test for students with hearing impairments. The data analysis used was the t-test between groups (independent t-test), which distinguished breaststroke swimming skills in students with hearing impairments between the experimental group and the control group. The experimental group was the group that was treated with the breaststroke swimming learning model based on Augmented Reality, while the control group was treated with the conventional learning model. The results showed that there was a significant difference in breaststroke swimming skills in students with hearing impairments between the experimental group and the control group, indicated by tcount 3.084 with p < 0.05. Judging from the average, the experimental group that was treated with the breaststroke swimming learning model based on Augmented Reality was 18.00 higher than the average in the control group was 13.36. This shows that the product of the breaststroke learning model based on Augmented Reality is more effective than the conventional breaststroke learning model. The conclusion in this study is that the product of the breaststroke learning model based on Augmented Reality has an influence on the development of breaststroke swimming skills in students with hearing impairments.

Project-Based Learning in Fostering Creative Thinking and Mathematical Problem-Solving Skills: Evidence from Primary Education in Indonesia

The interdependence between the Project-Based Learning (PjBL) Model and the growth and enhancement of Creative Thinking and Mathematical Problem Solving Skills in Elementary Schools is unquestionable nowadays. Prior studies have yet to discover concrete evidence regarding the interdependence being discussed. This study highlighted cognitive abilities related to creative thinking and mathematics problem-solving by implementing the Project-Based Learning Model. This research was a quasi-experiment with a pretest-posttest control group design involving 43 students in the sixth grade of two elementary schools; data was collected through test and classroom observation, and then the data was analyzed using Multivariate Analysis of Variance (MANOVA). Conversely, students exposed to project-based learning models exhibit higher skill levels in creative thinking and problem-solving than those instructed using conventional learning models. The project-based learning model significantly impacted elementary school children’s creative thinking and mathematics problem-solving skills. These findings suggest that the Project-Based Learning Model is acceptable for instructors seeking to foster creativity in teaching mathematics at the primary school level in Indonesia or other countries with comparable settings.

P2P credit risk management with KG-GNN: a knowledge graph and graph neural network-based approach

Credit risk management is crucial for the credit loan decision-making process on P2P platforms, essential for mitigating credit defaults. The rapid decline of China’s P2P platform market highlights the urgent need for Internet financial institutions to enhance their credit risk management strategies. Previous studies have applied machine learning to assess credit risk; however, their effectiveness is often hampered by a lack of consideration for semantic information and individual correlation. In response, we propose an innovative approach, KG-GNN, which integrates knowledge graph (KG) and graph neural network (GNN). KG-GNN leverages KG to encapsulate semantic information within complex categorical features and explore potential relationships between borrowers. Utilizing GNN, our framework extracts representation features from the KG to build comprehensive and accurate credit risk models. Our findings indicate that KG-GNN not only can predict credit risk more accurately than conventional machine learning models but also improves the stream model performance by over 20% through KG and GNN-based data augmentation techniques. By integrating KG with GNN, our approach enriches the methodologies for credit risk management and can be adapted to other data mining challenges that require processing complex semantic and relational information, thereby enhancing model learning capabilities.

Prediction of Cytochrome P450 Substrates Using the Explainable Multitask Deep Learning Models.

Cytochromes P450 (P450s or CYPs) are the most important phase I metabolic enzymes in the human body and are responsible for metabolizing ∼75% of the clinically used drugs. P450-mediated metabolism is also closely associated with the formation of toxic metabolites and drug-drug interactions. Therefore, it is of high importance to predict if a compound is the substrate of a given P450 in the early stage of drug development. In this study, we built the multitask learning models to simultaneously predict the substrates of five major drug-metabolizing P450 enzymes, namely, CYP3A4, 2C9, 2C19, 2D6, and 1A2, based on the collected substrate data sets. Compared to the single-task model and conventional machine learning models, the multitask fingerprints and graph neural networks model achieved superior performance with the average AUC values of 90.8% on the test set. Notably, the multitask model demonstrated its good performance on the small amount of substrate data sets such as CYP1A2, 2C9, and 2C19. In addition, the Shapley additive explanation and the attention mechanism were used to reveal specific substructures associated with P450 substrates, which were further confirmed and complemented by the substructure mining tool and the literature.

Realistic mathematics education: Mathematics e-modules in improving student learning outcomes

Realistic Mathematics Education (RME) has gradually influenced the development of mathematics learning worldwide, including in Indonesia. Through the implementation and development of online-based 21st-century education, RME is increasingly proving itself as an effective approach to improving the quality of mathematics teaching and student learning. Learning using the RME approach based on the e-module System aims to become a real learning experience for education and students in implementing mathematical concepts into web-based software systems. The type of research used is pseudo-experimental research with a population in this study of 3 schools, and the number of samples is 147 students. Hypothesis testing was carried out with the N-gain test on the experimental class and control class and the Linear Regression Analysis Test to determine the effect of Gender on improving the learning outcomes of the experimental class. The results of this study show that: 1) There is a difference in the improvement of learning outcomes of classes taught with the RME approach based on the e-module System with conventional learning models; 2) There is no significant difference in the improvement of learning outcomes of students taught using the RME approach based on the E-Module System based on gender characteristics; 3) Gender only contributes 0.1% to the improvement of learning outcomes while 99.9% of other factors.

Implementation of Levels of Inquiry Model in Improving Students' HOTS on Fluid Material

This research is motivated by the low level of students' higher order thinking skills (HOTS) in physics. The purpose of this study is to determine how the application of Levels of Inquiry (LoI) learning model in improving students' HOTS. The research design used was a quasi-experiment involving grade XI students in one of the high schools in Bandung City with a research sample of 35 students in the experimental class and 25 students in the control class. The instruments used in this study were higher order thinking skills test in the form of two-level multiple choices for fluid material and observation sheet of learning model implementation. Data analysis techniques include N-Gain, hypothesis testing and effect size. The results of the analysis showed that the application of the LoI learning model in the experimental class was in the excellent category and could improve students' HOTS in the moderate category with an N-gain value of 0.53. In the control class, the application of the conventional learning model can improve students' HOTS in the low category with an N-gain score of 0.18. In addition, the analysis of the Mann-Whitney test results showed that there was a significant difference in students' HOTS between the experimental group and the control group with an effect size in the very large category. This research proves that the application of the LoI learning model is effective in improving students' HOTS in fluid materials.

The Effect of Implementing Learning Models Contextual Teaching and Learning (CTL) toward Students’ Science Learning Outcomes of Class IV Cluster Ii at Donggo, Bima Regency

This research is an experimental study with a nonequivalent control group design which aims to determine: (1) An overview of the science learning model at SD Gugus II, Donggo District, Bima Regency (2) An overview of the application of science learning in SD Class IV Cluster II, Donggo District, Bima Regency using the model Contextual Teaching and Learning (CTL). (3) Overview of the learning outcomes of Class IV Science at Gugus II Elementary School, Donggo District, Bima Regency. (4) The effect of Contextual Teaching and Learning (CTL) model learning on Science Class IV Gusus II learning outcomes, Donggo District, Bima Regency. The population in this study were fourth grade students in Cluster II, Donggo District, Bima Regency, which consisted of SDN Inpres O'o, SDN O'o (Mpili), SDN Kamunti, SDN Inpres Kala, and SDN Manggekompo with a total of 74 students. While the samples used in this study were class IV SDN Kamunti as a control class, totaling 20 people and class IV at SDN Inpres O'o as an experimental class, totaling 23 people. The data from this study were obtained from learning outcomes tests with the subject matter of motion and style in the form of pretest and posttest as well as teacher and student activity observation sheets. The data analysis technique in this research is descriptive and inferential analysis. The results showed that: (1) The science learning model used in SD Gugus II, Donggo District, Bima Regency, was a lecture and group discussion learning model. (2) The application of Contextual Teaching and Learning (CTL) model learning in class IV SD Gugus II, Donggo District, Bima Regency has been in accordance with the implementation of learning where the teacher's teaching activities are in accordance with the set plan and the positive activity of students is high when compared to the previous implementation of the learning model Contextual Teaching and Learning (CTL). (3) The science learning outcomes of the control class (SDN Kamunti) students who applied the conventional learning model were lower than the results of the experimental class that applied the Contextual Teaching and Learning (CTL) learning model. (4) the application of the Contextual Teaching and Learning (CTL) learning model has a significant effect on the fourth grade science learning outcomes of SD Gugus II, Donggo District, Bima Regency. This is evidenced by the results of the Independent t-test. The t-test value is 5.344 and the t-table value is 2.019.

PMANet: Malicious URL detection via post-trained language model guided multi-level feature attention network

The expansion of the Internet has led to the widespread proliferation of malicious URLs, becoming a primary vector for cyber threats. Detecting malicious URLs is now essential for improving network security. The technological revolution spurred by pre-trained language models holds great promise for advancing the detection of malicious URLs. However, current research applying these models to URLs fails to address several crucial factors, including the lack of domain-specific adaptability, the omission of character-level information, and the neglect of both local detail extraction and low-order encoding information. In this paper, we propose PMANet, a pre-trained Language Model-Guided multi-level feature attention network, for addressing these issues. To facilitate a smooth transition of the pre-trained Transformer into the URL domain and to enable it to effectively capture information at both subword and character levels, we propose a post-training program that continues training the model on URLs using three self-supervised learning objectives: masked language model, noisy language model, and domain discrimination task. Subsequently, we develop a module to capture the output of each encoding layer, thus extracting hierarchical representations of URLs spanning from low-level to high-level. In addition, we propose a layer-wise attention mechanism that dynamically assigns weight coefficients to these feature layers based on their relevance. Finally, we apply spatial pyramid pooling to perform multi-scale down-sampling in order to obtain both local features and global context. PMANet achieves multifaceted integration in URL feature extraction, including capturing information at both the lexical and character levels, extracting features from low to high order, and discerning patterns at both global and local scales. We evaluate PMANet against challenging real-world scenarios, such as small-scale data, class imbalance, cross-dataset, adversarial attacks, and case studies on active malicious URLs. All experiments demonstrate that PMANet exhibits superiority over both the previous state-of-the-art pre-trained models and conventional deep learning models. Specifically, PMANet still achieves a 0.9941 AUC under adversarial attacks and correctly identifies all 20 actively malicious URLs in the case study. The code and data for our research are available at: https://github.com/Alixyvtte/Malicious-URL-Detection-PMANet.

Arabic dialect identification in social media: A hybrid model with transformer models and BiLSTM

Arabic Dialect Identification (ADI) is a challenging task in natural language processing applications due to its diversity and regional variations. Despite previous efforts, this task is still difficult. Therefore, this study aims to use transformers to address the issue of ADI on social media. A combination of two hybrid models is proposed in this study: one that combines Bidirectional Long Short-Term Memory (BiLSTM) with CAMeLBERT, and the second model that combines the BiLSTM model with AlBERT. In addition, a novel dataset comprising 121,289 user-generated comments from various social media network platforms and four major Arabic dialects (Egyptian, Jordanian, Gulf and Yemeni) was introduced. Several experiments have been conducted using conventional Machine Learning Classifiers (MLCs) and Deep Learning Models (DLMs) as baselines to measure the performance and effectiveness of the proposed models. In addition, binary classification is performed between two dialects to determine which are closest to each other. The performance of the model is measured using common metrics such as precision, recall, F-score and F-measure. Experiment results demonstrate the superior efficiency of the proposed hybrid models in ADI, CAMeLBERT with BiLSTM and ALBERT with BiLSTM, which both recorded an accuracy of 87.67 % and 86.51 %, respectively.

Thermokarst landslides susceptibility evaluation across the permafrost region of the central Qinghai-Tibet Plateau: Integrating a machine learning model with InSAR technology

Thermokarst landslides (TLs), which are made up of retrogressive thaw slumps (RTSs) and active-layer detachments slides (ALDSs), are quickly increasing in the central Qinghai-Tibet Plateau (QTP) permafrost area. TLs induce many environmental problems and threaten the safety of infrastructure. A landslide susceptibility map is crucial to prevent the negative impacts of these landslides. However, traditional thermokarst landslides susceptibility (TLS) evaluations do not consider real-time surface deformation information. Therefore, we propose a novel method that integrates a conventional machine learning model with ground surface deformation. Nine influencing factors, including slope, normalized difference vegetation index, elevation, precipitation, thawing degree days, soil content, active layer thickness, water content, and vegetation type were selected based on the q-index detector, and support vector machine was employed to obtain an initial model (IM). We obtained surface deformation in the study area using the enhanced Small Baseline Subset (SBAS) method. The accuracy of the InSAR results was validated through comparison with data from two field monitoring cross-sections. Subsequently, we established an integrated model (ITM) by combining the initial model with surface deformation using the contribution matrix, and confirmed the fusion model’s higher rationality and accuracy through comparison with the IM. Furthermore, we examined the impact of the quadtree segmentation method on atmospheric correction and validated the TLS results obtained using the ITM with high-resolution optical remote sensing imagery from GF6. Finally, the influencing factors and distribution characteristics of TLS was analyzed, and the results indicated that climatic conditions are the primary factors affecting the distribution of TLS.

Admission blood tests predicting survival of SARS-CoV-2 infected patients: a practical implementation of graph convolution network in imbalance dataset

BackgroundPredicting an individual’s risk of death from COVID-19 is essential for planning and optimising resources. However, since the real-world mortality rate is relatively low, particularly in places like Hong Kong, this makes building an accurate prediction model difficult due to the imbalanced nature of the dataset. This study introduces an innovative application of graph convolutional networks (GCNs) to predict COVID-19 patient survival using a highly imbalanced dataset. Unlike traditional models, GCNs leverage structural relationships within the data, enhancing predictive accuracy and robustness. By integrating demographic and laboratory data into a GCN framework, our approach addresses class imbalance and demonstrates significant improvements in prediction accuracy.MethodsThe cohort included all consecutive positive COVID-19 patients fulfilling study criteria admitted to 42 public hospitals in Hong Kong between January 23 and December 31, 2020 (n = 7,606). We proposed the population-based graph convolutional neural network (GCN) model which took blood test results, age and sex as inputs to predict the survival outcomes. Furthermore, we compared our proposed model to the Cox Proportional Hazard (CPH) model, conventional machine learning models, and oversampling machine learning models. Additionally, a subgroup analysis was performed on the test set in order to acquire a deeper understanding of the relationship between each patient node and its neighbours, revealing possible underlying causes of the inaccurate predictions.ResultsThe GCN model was the top-performing model, with an AUC of 0.944, considerably outperforming all other models (p < 0.05), including the oversampled CPH model (0.708), linear regression (0.877), Linear Discriminant Analysis (0.860), K-nearest neighbours (0.834), Gaussian predictor (0.745) and support vector machine (0.847). With Kaplan-Meier estimates, the GCN model demonstrated good discriminability between low- and high-risk individuals (p < 0.0001). Based on subanalysis using the weighted-in score, although the GCN model was able to discriminate well between different predicted groups, the separation was inadequate between false negative (FN) and true negative (TN) groups.ConclusionThe GCN model considerably outperformed all other machine learning methods and baseline CPH models. Thus, when applied to this imbalanced COVID survival dataset, adopting a population graph representation may be an approach to achieving good prediction.

MODEL DISCOVERY LEARNING BERBANTUAN GEOGEBRA UNTUK MENINGKATKAN KEMAMPUAN BERPIKIR KRITIS MATEMATIS DAN KEMANDIRIAN BELAJAR SISWA

Mathematics, as an important foundation in education, contributes to the advancement of science and technology and teaches students logical, analytical, and creative thinking. This is in accordance with the Merdeka Curriculum which emphasizes the development of 21st century skills and character development, including critical thinking and self-regulated learning. However, the low mathematical critical thinking skills and self-regulated learning in Indonesia is still a challenge. One of the factors that influence this is the use of inappropriate learning models. The Discovery Learning learning model supported by technology such as GeoGebra is a potential solution to improve both capabilities. The purpose of this study was to determine the increase in mathematical critical thinking skills and determine the increase in self-regulated learning of students who obtained the GeoGebra-assisted Discovery Learning model compared to students who obtained conventional learning models. This research method is a quasi-experiment with the nonequivalent pre-test post-test control group design. Sampling is done by purposive sampling technique. This research was carried out in one of the Junior High Schools in Serang City in the 2023/2024 school year, involving two classes of grade VII students as an experimental and control group. The instruments in this study consisted of tests of mathematical critical thinking skills and learning independence scales. The results showed that the improvement of mathematical critical thinking skills and increased self-regulated learning of students who applied the GeoGebra-assisted Discovery Learning model were better than students who applied conventional learning models

Hyper-local black carbon prediction by integrating land use variables with explainable machine learning model

This research introduces an innovative framework that effectively integrates the Land Use Regression (LUR) model with an explainable machine learning methodology, thereby enhancing the causal understanding of how land use variables impact Black Carbon (BC) concentrations in complex urban landscapes. Through the development and application of a LUR-based Bayesian Network (BN) model to detailed, mobile-based BC measurements, we have successfully decoded the causal relationships and conditional probabilities linking diverse urban land use variables to BC levels. The BN model demonstrates notable prediction accuracy on the test set, with an R2 of 0.59 and an NRMSE of 0.65. Among the varied land use variables analyzed, proximity to the port is identified as the dominant factor contributing to BC hotspots. Moreover, the synergistic effect of port proximity and truck routes emerges as a key driver of spatially clustered BC hotspots, a complex phenomenon not adequately captured by conventional machine learning and deep learning models. The insights revealed through this study are vital not just for enhancing air pollution prediction and exposure analysis but also for informing urban planning and policymaking, offering valuable, actionable insights.

PENERAPAN MODEL PEMBELAJARAN CREATIVE PROBLEM SOLVING (CPS) UNTUK MENINGKATKAN KEMAMPUAN PEMECAHAN MASALAH MATEMATIS DAN SELF-EFFICACY MATEMATIS SISWA

This research is motivated by a problem, namely the low ability to solve mathematical and self-efficacy mathematical problems where the learning process is still centered on the teacher. Thus given a solution using the Creative Problem Solving (CPS) learning model aims to determine the improvement of mathematical problem solving capabilities and mathematical self-efficacy students using the Creative Problem Solving (CPS) learning model is better than students taught with conventional learning models. The research approach is in the form of quantitative. The type of research used is Quasi experiment with non-equivalent Control Group Design research design. The population in the study was all students of class VII MTsS Jabal Nur, while the sample was class VII-1 as the experimental class and class VII-2 as the control class chosen with purposive sampling technique. The data collection techniques used are instruments for the ability to solve mathematical problems and non self-efficacy mathematical questionnaire tests are divided into pretest and post-test. Data is processed using SPSS software version 18. Analysis of data conducted for the mathematical problem solving capability test using non-parametric tests (Mann Whitney U) because there is data that is not normally distributed. Asymp sig value, (2-tailed) is 0,000 &lt;0.05, then is rejected and is accepted. Then for the analysis of students' mathematical self-efficacy questionnaires using the normality test, homogeneity. After the data is declared normal and homogeneous Furthermore, a hypothesis test with a significant value ˂ 0.05 is 0,000, then is rejected and is accepted. It can be concluded that the increase in mathematical problem solving capabilities and mathematical self-efficacy uses the Creative Problem Solving (CPS) learning model is better than students taught with conventional learning models in class VII MTsS Jabal Nur.