Generative Learning Model Research Articles

SVQ-MAE: an efficient speech pre-training framework with constrained computational resources

Self-supervised learning for speech pre-training models has achieved remarkable success in acquiring superior speech contextual representations by learning from unlabeled audio, excelling in numerous downstream speech tasks. However, the pre-training of these models necessitates significant computational resources and training duration, presenting a high barrier to entry into the realm of pre-training learning. In our efforts, by amalgamating the resource-efficient benefits of the generative learning model, Masked Auto Encoder, with the efficacy of the vector quantization method in discriminative learning, we introduce a novel pre-training framework: Speech Vector Quantization Masked Auto Encoder (SVQ-MAE). Distinct from the majority of SSL frameworks, which require simultaneous construction of speech contextual representations and mask reconstruction within an encoder-only module, we have exclusively designed a decoupled decoder for pre-training SVQ-MAE. This allows the additional decoupled decoder to undertake the mask reconstruction task solely, reducing the learning complexity of pretext tasks and enhancing the encoder’s efficiency in extracting speech contextual representations. Owing to this innovation, by using only 4 GPUs, SVQ-NAE can achieve high performance comparable to wav2vec 2.0, which requires 64 GPUs for training. In the Speech Processing Universal Performance Benchmark, SVQ-MAE surpasses wav2vec 2.0 in both keyword spotting and emotion recognition tasks. Furthermore, in cross-lingual ASR for Mandarin, upon fine-tuning on AISHELL-1, SVQ-MAE achieves a Character Error Rate of 4.09%, outperforming all supervised ASR models.

THE INFLUENCE OF A GENERATIVE LEARNING MODEL BASED ON WAVE MATERIAL COGNITIVE CONFLICT ON STUDENT LEARNING OUTCOMES AT SMAN 5 PAYAKUMBUH

Student learning outcomes in Physics subjects, especially in Wave material, are still low. This is caused by the learning model used in the Wave material not being being the process standards and characteristics of wave learning objectives. A generative learning model based on cognitive conflict is one approach to solving the aforementioned issue. The purpose of this study is to ascertain how student learning outcomes at SMAN 5 Payakumbuh are affected by the use of a generative learning model based on cognitive conflict in Wave content. The type of research is quasi-experimental research with a pretest-posttest Control Group Design. The participants in this study were all SMAN 5 Payakumbuh science class XI students. The information gathered reflects the attitudes, knowledge, and abilities of the pupils toward learning. Two averages are tested for equality as a data analysis technique. The study's findings demonstrate the notable distinctions between classrooms that employ the Problem-Based Learning model and those that employ the Generative learning model based on cognitive conflict. The average student learning outcome in classes that apply the Generative learning model is 83.14 and in classes that apply the Problem-Based Learning model is 77.26. The influence of the generative learning model based on cognitive conflict can be seen in the learning outcomes of students who have been analyzed and hypothesis tested. Based on the hypothesis test, the ttable is 2.006 and the tcount is 2.1. The condition for H0 to be rejected is if ttable < tcount. The tcount value is within the rejection of H0, so HI is accepted. Because all variables are controlled, except the learning model, it can be said that the application of the Generative learning model based on cognitive conflict in Wave material has a positive effect on student learning outcomes.

Integration of cognitive conflict in generative learning model to enhancing students’ creative thinking skills

In the complexity of the Fourth Industrial Revolution era, the importance of creative thinking is increasingly emphasized in the context of learning computing and algorithms. These skills are instrumental in inspiring innovative solutions, addressing complex challenges, and fostering the development of advanced technologies that characterize the transformative landscape of Industrial Revolution 4.0. This study aims to determine the effectiveness of the generative learning model based on cognitive conflict in improving the creative thinking skills (CTS) and learning outcomes of students in the computational physics and algorithms & programming courses. This research used mixed methods consisting of pretest-posttest control group design and snowballing technique. The research instruments consist of cognitive tests, psychomotor tests, affective tests, CTS tests, observation questionnaires, and interviews. The research sample consisted of 138 students taking computational physics and algorithms & programming courses. Quantitative data were analyzed using multivariate analysis of variance and qualitative data were analyzed using narrative analysis. The findings indicate that this model effectively improves students’ CTS and learning outcomes. Furthermore, the cognitive conflict aspect encourages students to be creative in analyzing and solving problems. This model has the potential to be used to optimize students’ potential in facing the demands of the fourth industrial revolution.

Continual Learning of Generative Models With Limited Data: From Wasserstein-1 Barycenter to Adaptive Coalescence.

Learning generative models is challenging for a network edge node with limited data and computing power. Since tasks in similar environments share a model similarity, it is plausible to leverage pretrained generative models from other edge nodes. Appealing to optimal transport theory tailored toward Wasserstein-1 generative adversarial networks (WGANs), this study aims to develop a framework that systematically optimizes continual learning of generative models using local data at the edge node while exploiting adaptive coalescence of pretrained generative models. Specifically, by treating the knowledge transfer from other nodes as Wasserstein balls centered around their pretrained models, continual learning of generative models is cast as a constrained optimization problem, which is further reduced to a Wasserstein-1 barycenter problem. A two-stage approach is devised accordingly: 1) the barycenters among the pretrained models are computed offline, where displacement interpolation is used as the theoretic foundation for finding adaptive barycenters via a "recursive" WGAN configuration and 2) the barycenter computed offline is used as metamodel initialization for continual learning, and then, fast adaptation is carried out to find the generative model using the local samples at the target edge node. Finally, a weight ternarization method, based on joint optimization of weights and threshold for quantization, is developed to compress the generative model further. Extensive experimental studies corroborate the effectiveness of the proposed framework.

NJmat: Data-Driven Machine Learning Interface to Accelerate Material Design.

Machine learning techniques have significantly transformed the way materials scientists conduct research. However, the widespread deployment of machine learning software in daily experimental and simulation research for materials and chemical design has been limited. This is partly due to the substantial time investment and learning curve associated with mastering the necessary codes and computational environments. In this paper, we introduce a user-friendly, data-driven machine learning interface featuring multiple "button-clicking" functionalities to streamline the design of materials and chemicals. This interface automates the processes of transforming materials and molecules, performing feature selection, constructing machine learning models, making virtual predictions, and visualizing results. Such automation accelerates materials prediction and analysis in the inverse design process, aligning with the time criteria outlined by the Materials Genome Initiative. With simple button clicks, researchers can build machine learning models and predict new materials once they have gathered experimental or simulation data. Beyond the ease of use, NJmat offers three additional features for data-driven materials design: (1) automatic feature generation for both inorganic materials (from chemical formulas) and organic molecules (from SMILES), (2) automatic generation of Shapley plots, and (3) automatic construction of "white-box" genetic models and decision trees to provide scientific insights. We present case studies on surface design for halide perovskite materials encompassing both inorganic and organic species. These case studies illustrate general machine learning models for virtual predictions as well as the automatic featurization and Shapley/genetic model construction capabilities. We anticipate that this software tool will expedite materials and molecular design within the scope of the Materials Genome Initiative, particularly benefiting experimentalists.

GPianoroll: a Deep Learning System with Human Feedback for Music Generation

We present a human-feedback-based evaluation pipeline for deep learning generative models, tested on music composition. Leveraging Bayesian Optimization and asking users to grade the generated songs, we efficiently find the point of a given model’s latent space that yields each volunteer’s favorite without retraining.

Efektivitas Model Pembelajaran Generatif: Pemahaman dan Keterlibatan Siswa dalam Pembelajaran Matematika

Enhancing the quality of mathematics education requires innovative and interactive teaching approaches. Generative Learning Model offers a framework that allows students to actively participate in the learning process, potentially improving their understanding of mathematical concepts. This research aims to conduct a systematic analysis of existing literature on the use of Generative Learning Model in mathematics education, focusing on effective strategies for developing interactive and effective learning tools. The research methodology involves developing a review protocol, conducting a comprehensive literature search through databases such as Google Scholar, and selecting studies using the PRISMA flowchart. The results of the analysis show that the Generative Learning Model can significantly improve mathematical concept understanding and student engagement. Effective strategies in developing learning tools based on the Generative Learning Model include the use of interactive tools and contextual approaches in teaching. The conclusions of this research confirm that the use of the Generative Learning Model contributes positively to the interactivity and effectiveness of mathematics learning, and provides recommendations for further research and practical application in educational contexts.

Efforts to Improve Mathematical Communication Skills in Mathematics Learning in Indonesia

This research aims to identify various models/methods/strategies and media that are effective in improving mathematical communication. This study uses the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) method with four stages, namely identification, scanning, eligibility, and inclusion. Harzing's Publish or Perish (PoP) is used to search for required research references. The research results found 30 publications regarding efforts to improve mathematical communication skills in schools and universities. Realistic Mathematic Education (RME), cooperative learning models, and the use of media are the methods most widely used in efforts to improve mathematical communication skills. Quantum learning models/methods, PBL, generative learning models, CBL, BBL, discovery, inquiry, and SSCS can also be chosen as alternatives in efforts to improve mathematical communication skills. Middle school level and geometry materials are most often used by researchers in efforts to improve mathematical communication. The results of this research can be a reference for teachers to choose the right model/method/strategy and media to improve mathematical communication skills in pupils and students.

The Effectiveness of the Generative Learning Model to Develop Audio Comprehension Skills and Language Thinking Skills among Primary School Students

The study aimed to verify the effectiveness of the generative learning model in developing audio comprehension skills and linguistic thinking; to achieve the objective of the study, the researcher used the semi-experimental approach, based on the design of the experimental and control groups, and the researcher also identified a list of audio comprehension skills and a list of linguistic thinking skills appropriate for sixth grade primary students, in the light of which an achievement test for audio comprehension skills and a linguistic thinking test were built, and two research materials were designed: The teacher's guide and the student's educational activities booklet; After the researcher verified the validity and stability of the two study tools, they were applied to the study sample consisting of (66) students. After processing the data statistically, the study found statistically significant differences at a level of significance less or equal to (0.05) between the average scores of the experimental group and the control group in the total dimensional test of audio comprehension skills and linguistic thinking, in favor of the experimental group, and the study also provided a number of recommendations and suggestions in light of the results of the study reached.

GenCoder: A Novel Convolutional Neural Network Based Autoencoder for Genomic Sequence Data Compression.

Revolutionary advances in DNA sequencing technologies fundamentally change the nature of genomics. Today's sequencing technologies have opened into an outburst in genomic data volume. These data can be used in various applications where long-term storage and analysis of genomic sequence data are required. Data-specific compression algorithms can effectively manage a large volume of data. In recent times, deep learning has achieved great success in many compression tools and is gradually being used in genomic sequence compression. Significantly, autoencoder has been applied in dimensionality reduction, compact representations of data, and generative model learning. It can use convolutional layers to learn essential features from input data, which is better for image and series data. Autoencoder reconstructs the input data with some loss of information. Since accuracy is critical in genomic data, compressed genomic data must be decompressed without any information loss. We introduce a new scheme to address the loss incurred in the decompressed data of the autoencoder. This paper proposes a novel algorithm called GenCoder for reference-free compression of genomic sequences using a convolutional autoencoder and regenerating the genomic sequences from a latent code produced by the autoencoder, and retrieving original data losslessly. Performance evaluation is conducted on various genomes and benchmarked datasets. The experimental results on the tested data demonstrate that the deep learning model used in the proposed compression algorithm generalizes well for genomic sequence data and achieves a compression gain of 27% over the best state-of-the-art method.

Generative Quantile Regression with Variability Penalty

Quantile regression and conditional density estimation can reveal structure that is missed by mean regression, such as multimodality and skewness. In this article, we introduce a deep learning generative model for joint quantile estimation called Penalized Generative Quantile Regression (PGQR). Our approach simultaneously generates samples from many random quantile levels, allowing us to infer the conditional distribution of a response variable given a set of covariates. Our method employs a novel variability penalty to avoid the problem of vanishing variability, or memorization, in deep generative models. Further, we introduce a new family of partial monotonic neural networks (PMNN) to circumvent the problem of crossing quantile curves. A major benefit of PGQR is that it can be fit using a single optimization, thus, bypassing the need to repeatedly train the model at multiple quantile levels or use computationally expensive cross-validation to tune the penalty parameter. We illustrate the efficacy of PGQR through extensive simulation studies and analysis of real datasets. Code to implement our method is available at https://github.com/shijiew97/PGQR.

Synthesising microstructures of a partially frozen salty sand using voxel-based 3D generative adversarial networks

Generating synthetic material microstructures is essential in the numerical modelling of geomaterials. The occurrence of permafrost and saline groundwater overlapping regions is crucial in a series of phenomena, such as carbon emissions and subgrade settlements. The microstructure of geomaterials in these regions is of particular complexity because of the multiphase nature with salty water and ice crystals. This complexity renders existing generative models ineffective in synthesising their microstructures. Traditional generative methods are limited in the sense that require prior knowledge of material descriptors. Recently, machine learning generative models achieved unprecedented levels of performance and realism, but still lack the means to assess posterior error. This work aims to bridge the gap between traditional methods and deep learning generative models by assessing posterior image quality in the latter. A 3D Generative Adversarial Network (GAN) model is trained with image samples from an X-ray CT of a partially frozen salty sand. The metrics retained to assess posterior quality are particle fabric (shape parameter and anisotropy) and homogenised elastic coefficients obtained with Finite Element Method (FEM) simulations. A hyperparametric study on batch size and latent dimension serves to select the best configuration based on particle fabric. FEM simulations determine the deviation in the generated images elastic coefficients being 7.55% on average with respect to real samples. With 803 voxels, generated images are the largest up-to-date in a three-phase material, allowing to reach REV criteria. Applications range from the generation of microscales in double-scale models to the calibration of image processing tools.

Mechanism for feature learning in neural networks and backpropagation-free machine learning models.

Understanding how neural networks learn features, or relevant patterns in data, for prediction is necessary for their reliable use in technological and scientific applications. In this work, we presented a unifying mathematical mechanism, known as average gradient outer product (AGOP), that characterized feature learning in neural networks. We provided empirical evidence that AGOP captured features learned by various neural network architectures, including transformer-based language models, convolutional networks, multilayer perceptrons, and recurrent neural networks. Moreover, we demonstrated that AGOP, which is backpropagation-free, enabled feature learning in machine learning models, such as kernel machines, that a priori could not identify task-specific features. Overall, we established a fundamental mechanism that captured feature learning in neural networks and enabled feature learning in general machine learning models.

Character-based physics learning module through generative learning model: Student conceptual understanding

One of the causes of students' low conceptual understanding is the lack of instructional materials with a composition that aids students in understanding concepts independently or during classroom learning. The aim of this research is to produce a character-based learning module, particularly emphasizing the virtues of honesty, valuing others' opinions, and caring, through a valid, practical, and effective generative learning model to enhance students' conceptual understanding. This study follows a research and development approach. The development design employs the ADDIE model, tested with 29 eleventh-grade science students at SMAN 6 Banjarmasin. Data collection involved validation sheets for the learning module, observations of lesson plan implementation, student response questionnaires, and learning outcome tests. The data were then analyzed qualitatively and quantitatively. The results show: (1) the developed character-based learning module with an average score of 2.9 is declared valid, (2) the module, assessed through lesson plan implementation and student response questionnaires, is considered practical, and (3) the developed character-based learning module is effective in improving students' conceptual understanding, yielding an N-gain of 0.65 categorized as moderate, with a noticeable difference in students' character achievement before and after the learning process. In conclusion, this study affirms that the character-based learning module with a generative learning model is deemed suitable for enhancing students' conceptual understanding.

Nonlinear system identification using modified variational autoencoders

This research proposes a methodology for identifying nonlinear systems using input/output data and deep learning generative models. Our framework integrates Variational Autoencoders (VAE) with Nonlinear Autoregressive with exogenous input (NARX) in a unified identification structure to address overfitting in nonlinear system identification using NARX structures. Specifically, we modify a variational autoencoder by replacing the decoder module with a NARX model using the latent space information captured from the VAE encoder module as one of the exogenous inputs. Following the training phase, the decoder module can be used as a nonlinear model of the system. We evaluate the efficacy of our approach by performing open-loop prediction tests on data from four nonlinear benchmark systems: Cascaded tanks, Gas furnace, Silverbox, and Wiener-Hammerstein. The proposed VAE-NARX method reported Root Mean Squared Error (RMSE) of 8.23×10−3, 16.69×10−3, 0.002×10−3 and 0.037×10−3 respectively. Our results demonstrate that our proposed method achieves similar and outperforms prediction performances to standard identification techniques and can enhance the performance of traditional nonlinear system identification methods based on multi-layer perceptron models.

A review of deep learning‐based approaches for deepfake content detection

AbstractRecent advancements in deep learning generative models have raised concerns as they can create highly convincing counterfeit images and videos. This poses a threat to people's integrity and can lead to social instability. To address this issue, there is a pressing need to develop new computational models that can efficiently detect forged content and alert users to potential image and video manipulations. This paper presents a comprehensive review of recent studies for deepfake content detection using deep learning‐based approaches. We aim to broaden the state‐of‐the‐art research by systematically reviewing the different categories of fake content detection. Furthermore, we report the advantages and drawbacks of the examined works, and prescribe several future directions towards the issues and shortcomings still unsolved on deepfake detection.

Analysis of the Implementation of the Generative Learning Model on the Achievement of Class X SMA Budisatrya Medan

The problem in this research is the low student learning outcomes. The aim of this research is to determine the influence of the Generative Learning Model on the Economics Learning Outcomes of Class This research was conducted in class X of Medan Budisatrya Private High School FY 2014/2015. The population in this study was all class X of Medan Budisatrya Private High School which consisted of 4 classes. This research sample consisted of X-3 as the experimental class and X-4 as the control class. The sampling technique used was simple random sampling. The instrument or data collection technique in this research is a test of student communication learning outcomes in the form of multiple choices of 20 items. Meanwhile, the data analysis technique used is the normality test, homogeneity test and to test the hypothesis the t test is used. The results of data analysis show that the average value of the experimental class is 64.75 with a standard deviation of 13.96, while the average value of the control class is 60.75 with a standard deviation of 13.99. Hypothesis testing is carried out using the t test statistical. From the results of the hypothesis calculation, the t_count is 5,750 and the t_table is 1,994. In other words, the hypothesis is accepted. From these results it can be concluded that there is an influence of the use of the Generative Learning Model on the Economic Learning Outcomes of Class

The Effect of a Strategy Based on the Generative Learning Model in Developing Social Problem Solving in the Subject of National and Civic Education for Basic Tenth Grade Students

This study aimed to investigate the effect of a strategy based on the generative education model in developing social problems solving in the subject of national and civic education among tenth grade students. The study followed the quasi-experimental methodology, and its members consisted of (148) male and female students from the basic tenth grade, with (76) students in Zaid bin Haritha school who were divided into two equally experimental and control groups, and (72) female students in Al-Qadisiyah school for girls were distributed into two groups, experimental and control group equally. To achieve the objectives of the study, a development test for social problems solving was prepared, which consisted of (25) multiple-choice items. The results showed that there was a statistically significant effect at the significance level (α = 0.05) on the experimental group that was studied using the generative education model for tenth graders of both sexes compared to the usual method. The study recommended the need to prepare an application guide for teachers of national and civic education for the tenth grade, in which it is clarified how to prepare and implement lessons using the generative learning model.

Learning Generative Models for Climbing Aircraft from Radar Data

Accurate trajectory prediction for climbing aircraft is hampered by the presence of epistemic uncertainties concerning aircraft operation, which can lead to significant misspecification between predicted and observed trajectories. This paper proposes a generative model for climbing aircraft in which the standard Base of Aircraft Data (BADA) model is enriched by a functional correction to the thrust that is learned from the data. The method offers three features: predictions of the arrival time with 66.3% less error when compared to BADA; generated trajectories that are realistic when compared to test data; and a means of computing confidence bounds for minimal computational cost.

Osl-dynamics, a toolbox for modeling fast dynamic brain activity.

Neural activity contains rich spatiotemporal structure that corresponds to cognition. This includes oscillatory bursting and dynamic activity that span across networks of brain regions, all of which can occur on timescales of tens of milliseconds. While these processes can be accessed through brain recordings and imaging, modeling them presents methodological challenges due to their fast and transient nature. Furthermore, the exact timing and duration of interesting cognitive events are often a priori unknown. Here, we present the OHBA Software Library Dynamics Toolbox (osl-dynamics), a Python-based package that can identify and describe recurrent dynamics in functional neuroimaging data on timescales as fast as tens of milliseconds. At its core are machine learning generative models that are able to adapt to the data and learn the timing, as well as the spatial and spectral characteristics, of brain activity with few assumptions. osl-dynamics incorporates state-of-the-art approaches that can be, and have been, used to elucidate brain dynamics in a wide range of data types, including magneto/electroencephalography, functional magnetic resonance imaging, invasive local field potential recordings, and electrocorticography. It also provides novel summary measures of brain dynamics that can be used to inform our understanding of cognition, behavior, and disease. We hope osl-dynamics will further our understanding of brain function, through its ability to enhance the modeling of fast dynamic processes.