Self-taught Learning Research Articles

MutexMatch: Semi-Supervised Learning With Mutex-Based Consistency Regularization.

The core issue in semi-supervised learning (SSL) lies in how to effectively leverage unlabeled data, whereas most existing methods tend to put a great emphasis on the utilization of high-confidence samples yet seldom fully explore the usage of low-confidence samples. In this article, we aim to utilize low-confidence samples in a novel way with our proposed mutex-based consistency regularization, namely MutexMatch. Specifically, the high-confidence samples are required to exactly predict "what it is" by the conventional true-positive classifier (TPC), while low-confidence samples are employed to achieve a simpler goal-to predict with ease "what it is not" by the true-negative classifier (TNC). In this sense, we not only mitigate the pseudo-labeling errors but also make full use of the low-confidence unlabeled data by the consistency of dissimilarity degree. MutexMatch achieves superior performance on multiple benchmark datasets, i.e., Canadian Institute for Advanced Research (CIFAR)-10, CIFAR-100, street view house numbers (SVHN), self-taught learning 10 (STL-10), and mini-ImageNet. More importantly, our method further shows superiority when the amount of labeled data is scarce, e.g., 92.23% accuracy with only 20 labeled data on CIFAR-10. Code has been released at https://github.com/NJUyued/MutexMatch4SSL.

Meta-DPSTL: meta learning-based differentially private self-taught learning

Meta-DPSTL: meta learning-based differentially private self-taught learning

Acquisition of Digital Audio Knowledge in the Studios of Senegalese Beatmakers.


 
 
 A field study carried out in the studios of some fifteen Senegalese beatmakers revealed the omnipresence of discourses linked to self-training to explain the ways in which they acquire the digital audio knowledge on which their creative practices are based. However, most of the time, this self-taught learning corresponds to situations where neither the learners nor the trainers recognize the formative nature of learning situations that actually take place. This article examines these processes in detail, in order to identify the types of learning that take place, and the types of knowledge that are transmitted. 
 
 

Deep self-taught learning framework for intrusion detection in cloud computing environment

<p>Cloud has become a target-rich environment for malicious attacks by cyber intruders. Security is a major concern and remains an obstacle to the adoption of cloud computing. The intrusion detection system (IDS) is regarded as defense-in-depth. Unfortunately, most machine learning approaches designed for cloud intrusion detection require large amounts of labeled attack samples, but in real practice, they are limited. Therefore, the key impetus of this work is to introduce self-taught learning (STL) combining stacked sparse autoencoder (SSAE) with long short-term memory (LSTM) as a candidate solution to learn the robust feature representation and efficiently improve the performance of IDS with respect to false alarm rate (FAR) and detection rate (DR). Accordingly, the proposed approach as a first step employs SSAE to achieve dimensional reduction by learning the discriminative features from network traffic. The approach adopts LSTM to recognize the intrusion with the features encoded by SSAE. To evaluate the detective performance of our model, a comprehensive set of experiments are conducted on NSL-KDD. Also, ablation experiments are conducted to show the contribution of each component of our approach. Further, the comparative analysis shows the efficacy of our approach against the existing approaches with an accuracy of 86.31%.</p>

Self-Supervised Self-Organizing Clustering Network: A Novel Unsupervised Representation Learning Method.

Deep learning-based clustering methods usually regard feature extraction and feature clustering as two independent steps. In this way, the features of all images need to be extracted before feature clustering, which consumes a lot of calculation. Inspired by the self-organizing map network, a self-supervised self-organizing clustering network ( [Formula: see text]OCNet) is proposed to jointly learn feature extraction and feature clustering, thus realizing a single-stage clustering method. In order to achieve joint learning, we propose a self-organizing clustering header (SOCH), which takes the weight of the self-organizing layer as the cluster centers, and the output of the self-organizing layer as the similarities between the feature and the cluster centers. In order to optimize our network, we first convert the similarities into probabilities which represents a soft cluster assignment, and then we obtain a target for self-supervised learning by transforming the soft cluster assignment into a hard cluster assignment, and finally we jointly optimize backbone and SOCH. By setting different feature dimensions, a Multilayer SOCHs strategy is further proposed by cascading SOCHs. This strategy achieves clustering features in multiple clustering spaces. [Formula: see text]OCNet is evaluated on widely used image classification benchmarks such as Canadian Institute For Advanced Research (CIFAR)-10, CIFAR-100, Self-Taught Learning (STL)-10, and Tiny ImageNet. Experimental results show that our method significant improvement over other related methods. The visualization of features and images shows that our method can achieve good clustering results.

Adapting the Education and Training System to Technological Evolution

Teachers have faced the challenge of adjusting their pedagogical approaches and teaching methods to help students with disabilities adapt to the distance education context, often without specific training or adequate resources. To address these challenges, the "Do IT Yourself" project aims to increase the involvement of students with disabilities in digital education. Its primary focus is promoting "innovative practices in the digital age" with a two-fold approach: encouraging social inclusion and using technological means to integrate vulnerable student groups. This study aims to highlight teachers' digitisation needs and the digital tools they use in the classroom. The data obtained supporting this quantitative study are part of the Erasmus Plus "Do It Yourself" project, of which Romania is a part, together with Portugal, Italy, Finland and Slovenia. The main objective of this study is to obtain data on the digitisation of education in Romania. The present study applies a research method that focuses on completing structured questions. In this research, 79 participants were involved, representing both pre-university teachers and school principals. The data collection process was carried out using a questionnaire consisting of 12 questions to evaluate the type of digital resources used in the teaching environment, the frequency of their use, and the level of training of the participants regarding the training courses for the integration of the tools digital or self-taught learning.

Experiencias del profesional de enfermería durante el cuidado al paciente crítico: lecciones autoaprendidas

Patient care should be provided in a favorable environment, considering that it is an attitude endowed with understanding, compassion, awareness and competence, taking into account the essence of the profession. Objective: to analyze the nursing professional's experiences during patient care in critical care units. The approach was qualitative, with descriptive design and a phenomenological approach. Ten participants were considered through non-probabilistic sampling and theoretical saturation. The instrument was the semi-structured interview. The data analysis was carried out through Colaizzi's method. Six categories emerged as a result of the research: experiences of the nursing professional during patient care in critical care units; elements that stand out in the affairs of the nursing professional during patient care in critical care units; repercussions of self-taught learning in the nursing professional; humanized comprehensive care; knowledge of the nursing staff on the management of the critical patient; and increased patient-family-nurse bonding. Final considerations According to the findings, in the nursing professional's experiences, deficiencies are perceived in the communication provided to the patient and family with whom the nursing staff should establish an interpersonal bond. Work overload is an element that influences the quality of care of patients admitted to the ICU. Keywords: care; humanization; nurse; patient.

Management of Self-Taught Learning for Non-Religious Background Students toward Culture Shock at State Islamic University

The purpose of this study was to explore the challenges of students from non-religious backgrounds when facing culture shock and highlight the management of self-taught learning carried out by students of one of the Islamic tertiary institutions in Bandung. This article used qualitative research method, a case study. Primary data source came from interviews with eight participants. The data analysis technique used was data condensation, data presentation, and drawing conclusions. The results showed that students with non-religious backgrounds who majored in religion at Islamic tertiary institutions experienced culture shock when participating in learning. The culture shock was in the form of being left behind in understanding the core theories of the department, due to their educational background coming from a public school. It required the students to start from zero compared to students with religious education backgrounds (Madrasah Aliyah and pesantren) who could directly participate in learning and deepen the knowledge. Based on the results of interviews with the respondents, the most common self-taught learning they did were study in their free time, study hard and study outside class hours, by reading books from various literacy.

Electronic-Based Marketing Strategies in Building Competitive Advantages in Silk Weaving Small Businesses in Wajo District

This study aims to determine the internal and external constraints on silk artisans. This research uses a qualitative descriptive method. The data used in this study are primary and secondary data, this data was obtained by researchers through observation, interviews, and documentation. The results of this study indicate that the electronic-based marketing strategy in building competitive advantages in small silk artisan businesses in Wajo Regency is by conducting a product marketing mix, determining the selling price of products based on the price of Production Operating Costs, a strategic and easily accessible business place, and promotions carried out, namely offline and online promotions. While the obstacles experienced by silk entrepreneurs in carrying out marketing strategies through Facebook social media are internal and external obstacles. The solution to overcome internal obstacles is to carry out an order system, self-taught learning through youtube media. for external obstacles is to provide affordable prices by using quality materials, be more creative and provide the best service, provide rich writing bounty marks on every picture posted, provide clear information, especially the prices offered.

Indonesian Tour Guide's Bilingual Experiences in Malang: A Narrative Inquiry Study

`The purpose of this study was to explore the experiences of a bilingual Indonesian tour guide in guiding foreigners coming to Malang, Indonesia. More specifically, this study intended to explore the challenges and strategies in learning and applying two languages. Narrative inquiry is employed in this study to examine the participant’s experiences before and after learning and practicing two languages. This study used a qualitative research design in which the data is described in the form of a description. The research instrument was in the form of open-ended interviews. The participant of this research is a senior tour guide who lives in Malang. The findings of this narrative inquiry indicated that less practice could hinder the participant in applying the two languages and hinder the participant from communicating with foreigners. To cope with the challenges, the participant applied communication strategies by self-taught learning and practice. The study suggests it is crucial to always use and practice the languages, especially for native speakers, to master them.
 Keywords: bilingualism, learning strategies, narrative inquiry

Rancang Bangun Media Pembelajaran Teknologi WAN Berbasis Android

The impact of the Covid-19 pandemic in the education sector is getting children used to using smartphones. Smartphones during a pandemic helped teachers deliver learning materials and became one of the learning media during the pandemic. Smartphones help teachers and students in the learning process during a pandemic. Learning media is a tool in the learning process. Learning helps teachers convey information in more interesting way. Current technological developments really help many people at work. In the education sector, technology can help in making learning media more interesting, for example by using the PowerPoint application in learning. This learning is made with an Android-based platform. Android-based learning media is learning media in the form of applications that are installed on smartphones and can be accessed directly by students. This Android-based learning media was developed using a prototyping model approach with the Adobe CS6 application. one of the functions of smartphones for teachers and students is as a learning medium. The benefits of Android-based learning media are that students can access it anywhere and anytime and support self-taught learning.

On the benefits of self-taught learning for brain decoding.

We study the benefits of using a large public neuroimaging database composed of functional magnetic resonance imaging (fMRI) statistic maps, in a self-taught learning framework, for improving brain decoding on new tasks. First, we leverage the NeuroVault database to train, on a selection of relevant statistic maps, a convolutional autoencoder to reconstruct these maps. Then, we use this trained encoder to initialize a supervised convolutional neural network to classify tasks or cognitive processes of unseen statistic maps from large collections of the NeuroVault database. We show that such a self-taught learning process always improves the performance of the classifiers, but the magnitude of the benefits strongly depends on the number of samples available both for pretraining and fine-tuning the models and on the complexity of the targeted downstream task. The pretrained model improves the classification performance and displays more generalizable features, less sensitive to individual differences.

“Discovering a glycoprotein: The case of the H,K-ATPase”. An Online Game for Improvement of Reading Skills in a Course of Biological Chemistry

During the COVID-19 pandemic we developed an activity that encourages engagement with scientific reading and helps students to improve their reading skills as a complement to the glycobiology section of an undergraduate course of biological chemistry. Using Genially, an online platform for gamification and self-taught learning, we designed the activity named “Discovering a glycoprotein: The case of the H,K-ATPase” based on the characterization of the β-subunit of the gastric H,K-ATPase, a glycoprotein discovered in the 1960s. While testing users’ knowledge in glycobiology, the game invites them to follow the research steps throughout four missions, where the experimental details and their results are presented. After solving the missions, participants were invited to answer a questionnaire through Google forms. Our results showed that the game is useful for students as it introduces an innovative way of analyzing a scientific article different from the conventional activities that students typically face without teacher assistance and in many cases do not complete. We expect that our pilot-study will be an additional boost for the implementation of gamification in the science classroom, and in turn, serve as an example of the importance of gamification in future teacher training as an assisting tool to traditional education methods.

STAR-Lite: A light-weight scalable self-taught learning framework for older adults’ activity recognition

Recognizing activities for older adults is challenging as we observe a variety of activity patterns caused due to aging (e.g., limited dexterity, limb control, slower response time) or/and underlying health conditions (e.g., dementia). However, existing literature with deep learning methods has successfully recognized activities when the dataset contains high-quality annotations and is captured in a controlled environment. On the contrary, data captured in a real-world environment, especially with older adults exhibiting memory-related symptoms, varying psychological and mental health status, reliance on caregivers to perform daily activities, and unavailability of domain-specific annotators, makes obtaining quality data with annotations challenging; leaving us with limited labeled data and abundant unlabeled data. In this paper, we hypothesize that projecting the labeled data representations comprising a specific set of activities onto a new representation space characterized by the unlabeled data comprising activities beyond the limited activities in the labeled dataset would help us rely less on the annotated data to improve activity detection performance. Motivated by this, we propose STAR-Lite, a self-taught learning framework that involves a pre-training framework to prepare the new representation space considering activities beyond the initial labels in the labeled dataset. STAR-Lite projects the labeled data representations on the new representation space characterized by unlabeled data labels and learns higher-level representations of the labeled dataset while optimizing inter- and intra- class distances without explicitly using a computation hungry similarity-based approach. We demonstrate that our proposed approach, STAR-Lite (a) improves activity recognition performance in a supervised setting and (b) is feasible for real-world deployment. To enhance the feasibility of deploying STAR-Lite on devices with limited memory resources, we explore model compression techniques such as pruning and quantization and propose a novel layer-wise pruning-rate optimization technique that effectively compresses the network while preserving the model performance. The evaluation was performed using the Alzheimer’s Activity Recognition dataset (AAR) captured from 25 individuals living in a retirement community center with IRB approval (#Y18NR12035) using an in-house SenseBox infrastructure while concurrently assessing the clinical evaluation of the participants for dementia, and independent living. Our extensive evaluation reveals that STAR-Lite can detect activities with an F1-score of 85.12% despite 62% reduction in model size and 5% improvement of execution time on a resource constrained device.

Tensor Dictionary Self-Taught Learning Classification Method for Hyperspectral Image

Precise object classification based on Hyperspectral imagery with limited training data presents a challenging task. We propose a tensor-based dictionary self-taught learning (TDSL) classification method to provide some insight into these challenges. The idea of TDSL is to utilize a small amount of unlabeled data to improve the supervised classification. The TDSL trains tensor feature extractors from unlabeled data, extracts joint spectral-spatial tensor features and performs classification on the labeled data set. These two data sets can be gathered over different scenes even by different sensors. Therefore, TDSL can complete cross-scene and cross-sensor classification tasks. For training tensor feature extractors on unlabeled data, we propose a sparse tensor-based dictionary learning algorithm for three-dimensional samples. In the algorithm, we initialize dictionaries using Tucker decomposition and update these dictionaries based on the K higher-order singular value decomposition. These dictionaries are feature extractors, which are used to extract sparse joint spectral-spatial tensor features on the labeled data set. To provide classification results, the support vector machine as the classifier is applied to the tensor features. The TDSL with the majority vote (TDSLMV) can reduce the misclassified pixels in homogenous regions and at the edges of different homogenous regions, which further refines the classification. The proposed methods are evaluated on Indian Pines, Pavia University, and Houston2013 datasets. The classification results show that TDSLMV achieves as high as 99.13%, 99.28%, and 99.76% accuracies, respectively. Compared with several state-of-the-art methods, the classification accuracies of the proposed methods are improved by at least 2.5%.

The Pattern of Development of Competence, Commitment, and Motivation of Teachers in Pesantren

This study aims to analyze the development of competence, commitment, and motivation of teachers in the Mu'adah Islamic boarding school. This study used a qualitative method with a multi-case type and was conducted at the Al-Amin Sumenep and Sidogiri Islamic boarding schools, Pasuruan. Data collection techniques used the instruments of the interview, observation, and documentation. Several activities in analyzing data include; data condensation, data presentation, and conclusion drawing. The results of this study found that the development of teacher competence at the Mu' Islamic boarding school is to apply a pattern based on spirituality and professionalism through integrated, holistic, and continuous coaching and training strategies, as well as to cultivate self-taught learning. The development of teacher commitment is based on the principle of devotion because Kyai, coaches, teachers, and students live together and are supported by a religiously nuanced Islamic boarding school environment. The development of teacher motivation is based on the values of sincerity and blessing combined with modern professional management.

Lessons learned from implementing a self-taught learning resource on sustainable healthcare for the UK medical curriculum

Lessons learned from implementing a self-taught learning resource on sustainable healthcare for the UK medical curriculum

THE STOCK PURCHASE STRATEGY DURING THE COVID-19 PANDEMIC

Since the COVID-19 pandemic throughout 2020, the Composite Stock Price Index (JCI) has decreased by minus 18.3% (idx.channel.okezone.com) and this is almost completely evenly distributed across several trading floors in the world. The urgency (priority) of this research is to be able to add a new perspective on the business environment analysis model as an investment strategy in determining and determining stock purchase decisions appropriately while considering internal and external business environmental factors that are able to dynamically influence investors' perceptions in minimizing losses on their portfolio of shares. This research method is a qualitative descriptive method. The key informants in this study were the manager of mutual securities in Bandung, the manager of UOB Kay Han Securities in the Bandung branch, and the manager of the securities in the Bandung branch. The Indonesia Stock Exchange (IDX) noted that the stock investor base, which incidentally is the millennial generation, has reached 60% of the total stock investors on the IDX. However, the results of a New Harris poll through the Stash application as quoted from the Financial Services Authority (OJK), almost 80% of young people do not even look at investing in the stock market. As many as 34% of the younger generation stated that they were reluctant to invest in stocks because the working system in the stock market was too difficult to learn or understand. Yes, you know, there are many terms in the capital market, especially stocks, such as bullish, bearish, and many others. Based on an interview with the Manager of Danareksa Sekuritas Branch Bandung as an informant, starting a stock investment is very easy. Capital for investment is currently starting from Rp. 100 thousand. No need to deposit millions of rupiah, just learn stock on the website, join trainings or enter capital market school. Self-taught learning can also be done by asking friends who have already entered the capital market or reading books and finding out on the internet. For stock investment, you must understand the basic knowledge first. Starting from learning to read the movement of the Composite Stock Price Index (JCI), when is the right time to buy or sell shares, to learning strategies and mitigating the risk of loss. To analyze and choose a strategy for buying shares during the COVID-19 pandemic, there are two basic approaches, namely : Fundamental Analysis Strategy and Technical Analysis Strategy.

STL-Detector: Detecting City-Wide Ride-Sharing Cars via Self-Taught Learning

Ride-sharing cars are private vehicles held by individuals or provided by ride-hailing companies for designated drivers to offer taxi-like services. Recently, various ride-sharing cars have emerged around the city with the popularity of online ride-hailing services. Identifying them is the critical task of transportation management. However, less work focuses on this issue due to the lack of city-wide private vehicles&#x2019; trajectory data and labeled ride-sharing cars. Fortunately, data collected by advanced sensing technology, such as electronic registration identification (ERI) of the motor vehicle data collected by radio-frequency identification (RFID) technology, provide us with an opportunity to detect ride-sharing cars from a data-driven aspect. This article proposes detecting ride-sharing cars via self-taught learning (STL) using ERI data, named STL-detector, which is accurate with very little labeled information. In detail, STL-detector consists of two components. In the <i>unsupervised feature learning</i> component, we construct a 3-D convolutional neural networks (3-D-CNN) autoencoder trained with an amount of unlabeled data, which forms a succinct high-level input representation and significantly improve detection performance. In the <i>supervised classification</i> component, we utilize the random forest (RF) as the classifier, which is trained on very little labeled data, to detect ride-sharing cars/others. The experimental results demonstrate that our STL-detector model can detect ride-sharing cars with better performance compared with other baselines on a set of train and test samples. Furthermore, we apply our model to a real-world scenario to detect ride-sharing cars and conduct a comparative analysis on the behavior of detected ride-sharing cars and taxis.

Latent-Insensitive Autoencoders for Anomaly Detection

Reconstruction-based approaches to anomaly detection tend to fall short when applied to complex datasets with target classes that possess high inter-class variance. Similar to the idea of self-taught learning used in transfer learning, many domains are rich with similar unlabeled datasets that could be leveraged as a proxy for out-of-distribution samples. In this paper we introduce the latent-insensitive autoencoder (LIS-AE) where unlabeled data from a similar domain are utilized as negative examples to shape the latent layer (bottleneck) of a regular autoencoder such that it is only capable of reconstructing one task. We provide theoretical justification for the proposed training process and loss functions along with an extensive ablation study highlighting important aspects of our model. We test our model in multiple anomaly detection settings presenting quantitative and qualitative analysis showcasing the significant performance improvement of our model for anomaly detection tasks.