scholarly journals Meta-Strategy for Learning Tuning Parameters with Guarantees

Entropy ◽  
2021 ◽  
Vol 23 (10) ◽  
pp. 1257
Author(s):  
Dimitri Meunier ◽  
Pierre Alquier
Keyword(s):  
Online Learning ◽  
Learning Rate ◽  
Gradient Algorithm ◽  
Step Size ◽  
Learning Methods ◽  
Tuning Parameters ◽  
Meta Learning ◽  
Regret Bound ◽  
Exponentially Weighted

Online learning methods, similar to the online gradient algorithm (OGA) and exponentially weighted aggregation (EWA), often depend on tuning parameters that are difficult to set in practice. We consider an online meta-learning scenario, and we propose a meta-strategy to learn these parameters from past tasks. Our strategy is based on the minimization of a regret bound. It allows us to learn the initialization and the step size in OGA with guarantees. It also allows us to learn the prior or the learning rate in EWA. We provide a regret analysis of the strategy. It allows to identify settings where meta-learning indeed improves on learning each task in isolation.

Limited Stochastic Meta-Descent for Kernel-Based Online Learning

2009 ◽  
Vol 21 (9) ◽  
pp. 2667-2686 ◽  
Author(s):  
Wenwu He
Keyword(s):  
Online Learning ◽  
Adaptive Learning ◽  
Learning Rate ◽  
Support Vector ◽  
Data Sets ◽  
Step Size ◽  
Adaptive Learning Rate ◽  
Practical Performance ◽  
Convergence Proofs ◽  
Theoretical Results

To improve the single-run performance of online learning and reinforce its stability, we consider online learning with limited adaptive learning rate in this letter. The letter extends convergence proofs for NORMA to a range of step sizes, then employs support vector learning with stochastic meta-descent (SVMD) limited to that range for step size adaptation, so as to obtain an online kernel algorithm that combines theoretical convergence guarantees with good practical performance. Experiments on different data sets corroborate theoretical results well and show that our method is another promising way for online learning.

A new adaptive variable step size natural gradient BSS algorithm

2021 ◽  
pp. 1-12
Author(s):  
Junqing Ji ◽  
Xiaojia Kong ◽  
Yajing Zhang ◽  
Tongle Xu ◽  
Jing Zhang
Keyword(s):  
Signal To Noise Ratio ◽  
Wavelet Coefficient ◽  
Gradient Algorithm ◽  
Variable Step Size ◽  
Natural Gradient ◽  
Threshold Method ◽  
Step Size ◽  
Morphological Component Analysis ◽  
Variable Step ◽  
Separation Degree

The traditional blind source separation (BSS) algorithm is mainly used to deal with signal separation under the noiseless model, but it does not apply to data with the low signal to noise ratio (SNR). To solve the problem, an adaptive variable step size natural gradient BSS algorithm based on an improved wavelet threshold is proposed in this paper. Firstly, an improved wavelet threshold method is used to reduce the noise of the signal. Secondly, the wavelet coefficient layer with obvious periodicity is denoised using a morphological component analysis (MCA) algorithm, and the processed wavelet coefficients are recombined to obtain the ideal model. Thirdly, the recombined signal is pre-whitened, and a new separation matrix update formula of natural gradient algorithm is constructed by defining a new separation degree estimation function. Finally, the adaptive variable step size natural gradient blind source algorithm is used to separate the noise reduction signal. The results show that the algorithm can not only adaptively adjust the step size according to different signals, but also improve the convergence speed, stability and separation accuracy.

Establishing Millennial Character Behaviors in Online Studies at The Covid-19 Pandemic in Indonesia

2021 ◽  
Author(s):  
Mochamad Zaenal Fanani
Keyword(s):  
Online Learning ◽  
Literature Review ◽  
Character Education ◽  
Millennial Generation ◽  
Learning System ◽  
Specific Information ◽  
Theoretical Frameworks ◽  
Learning Methods ◽  
Optimal Learning ◽  
Research Activities

This covid pandemic has changed everything 180 degrees, especially in the world of education. Students who are accustomed to face-to-face learning are required to move in the process of learning methods. The learning method implemented during this pandemic uses online learning. This measure was taken to maintain the quality of education and to prevent the spread of Covid-19. With online meetings, there will be many changes in learning methods to align the essence of learning appropriately and effectively because when compared to previous learning methods, it is very different. This article will highlight various online learning methods to always build the behavior characteristics of the millennial generation for the better. The question that arises from this topic is what is the right method of managing character education in a pandemic situation like this? The research aims to identify what problems occur in online learning and provide effective methods for the learning process in Indonesia. The writing of this article uses the literature review method, a description of the theory, findings, and other research materials obtained from reference materials to be used as the basis for research activities to develop a clear frame of mind from the formulation of the problem to be studied. The data are from the latest research, namely from 2019 to 2021 published through Google Scholar, totaling 12 articles, and data were analyzed qualitatively using available theoretical frameworks. The results of this study show the behavioral process of the millennial generation in building character amid online learning, and to create optimal learning, distance-learning or online strategies are carried out using methods to maximize the optimal learning system. From the research, literature review methods proved to be effective to study this topic while Covid-19 is happening to keep safety measures, but it does not give many details and specific information well. This method is recommended for researchers who work from home without having to go to the field to find information and avoid exposure to COVID-19.

Convergence analysis of online learning algorithm with two-stage step size

2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Weilin Nie ◽  
Cheng Wang
Keyword(s):  
Online Learning ◽  
Optimization Problems ◽  
Learning Algorithm ◽  
Computational Cost ◽  
Reproducing Kernels ◽  
Step Size ◽  
Two Stage ◽  
Logarithmic Term ◽  
Convergence Performance ◽  
Online Learning Algorithm

Abstract Online learning is a classical algorithm for optimization problems. Due to its low computational cost, it has been widely used in many aspects of machine learning and statistical learning. Its convergence performance depends heavily on the step size. In this paper, a two-stage step size is proposed for the unregularized online learning algorithm, based on reproducing Kernels. Theoretically, we prove that, such an algorithm can achieve a nearly min–max convergence rate, up to some logarithmic term, without any capacity condition.

Deep Learning Based Classification of Time Series of Chaotic Systems over Graphic Images

2021 ◽  
Author(s):  
Süleyman UZUN ◽  
Sezgin KAÇAR ◽  
Burak ARICIOĞLU
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Transfer Learning ◽  
Chaotic Systems ◽  
Initial Conditions ◽  
Step Size ◽  
Learning Methods ◽  
Data Set ◽  
Graphic Images ◽  
Different Chaotic Systems

Abstract In this study, for the first time in the literature, identification of different chaotic systems by classifying graphic images of their time series with deep learning methods is aimed. For this purpose, a data set is generated that consists of the graphic images of time series of the most known three chaotic systems: Lorenz, Chen, and Rossler systems. The time series are obtained for different parameter values, initial conditions, step size and time lengths. After generating the data set, a high-accuracy classification is performed by using transfer learning method. In the study, the most accepted deep learning models of the transfer learning methods are employed. These models are SqueezeNet, VGG-19, AlexNet, ResNet50, ResNet101, DenseNet201, ShuffleNet and GoogLeNet. As a result of the study, classification accuracy is found between 96% and 97% depending on the problem. Thus, this study makes association of real time random signals with a mathematical system possible.

scholarly journals Pendekatan Realita Perkuliahan Pada Media Online

2017 ◽  
Vol 15 (30) ◽  
pp. 1
Author(s):  
Surya Eka Priyatna
Keyword(s):  
Online Learning ◽  
Conventional Method ◽  
Online Media ◽  
Positive Contribution ◽  
Learning Methods ◽  
Lecture Method

In fact, the conventional method is still widely used in the lecture. The lecture method is one of the most popular among lecturers. This method is still regarded as an effective method for delivery of material to the students. The development of online learning provide a positive contribution to the development of learning methods. To achieve optimal lecture, then searched the conventional lecture approach reality in online media. Pada kenyataannya metode konvensional dalam perkuliahan masih sering digunakan. Metode ceramah merupakan salah satu yang paling populer di kalangan dosen. Metode ini masih dipandang sebagai metode yang efektif untuk penyampaian materi kepada mahasiswa. Berkembangnya pembelajaran online memberikan kontribusi yang positif pada perkembangan metode pembelajaran. Untuk mencapai tujuan perkuliahan yang optimal, maka dicari pendekatan realita perkuliahan konvensional pada media online.

scholarly journals Dampak Metode Pembelajaran Dari bagi Ketuntasan Belajar Siswa

2021 ◽  
Vol 1 (2) ◽  
pp. 117-129
Author(s):  
Franty Faldy Palempung ◽  
Ferry J N Sumual
Keyword(s):  
Online Learning ◽  
Qualitative Methods ◽  
Student Learning ◽  
Learning Process ◽  
Literature Study ◽  
Online Learners ◽  
Learning Methods ◽  
Knowledge And Skills ◽  
Study Approach ◽  
The Impact

­Abstrak: Tulisan ini secara spesifik memaparkan dampak metode pembelajaran daring bagi ketuntasan belajar siswa. Peristiwa Covid-19 yang terjadi awal tahun 2020 hingga sampai sekarang, masih menyebakan kesulitan bagi semua element. Imbas dari pandemi ini di sektor Pendidikan mengharuskan pembelajaran online dilaksanakan. Akibat dari penerapan pembelajaran daring ini, masih menyebabkan kesulitan bagi sebagai indvidu bahkan institusi karena masih belum lengkapnya pra-sarana dalam kegiatan pembelajara daring. Topik ini ditulis dengan menggunakan metode kualitatif deskriftif dengan pendekatan studi literatur. Hasil pada uraian ini mengemukakan bahwa pengertian ketuntasan belajar menurut Permendikbud adalah pencapaian kompetensi sikap, pengetahuan, dan keterampilan meliputi ketuntasan penguasaan substansi dan ketuntasan belajar dalam konteks kurun waktu belajar. Ketuntasan belajar peserta didik merupakan komponen keluaran yang diperoleh dari hasil suatu proses pembelajaran yang didukung oleh orang tua, guru dan lingkungan. Berhasil tidaknya pembelajaran daring bagi ketuntasan pembelajaran, diperlukan kerja sama semua komponen Pendidikan itu sendiri.Abstract: This paper specifically describes the impact of online learning methods on the completion of student learning. The Covid-19 event that occurred in early 2020 until now, still makes it difficult for all elements. The impact of this pandemic in the Education sector requires that online learning be implemented. As a result of the application of online learning in, still causes difficulties for as an individual even an institution because it is still incomplete pre-facilities in the activities of online learners. This topic is written using qualitative methods with a literature study approach. The results in this description suggest that the understanding of the completion of learning according to Permendikbud is the achievement of attitude competence. Knowledge, and skills include the completion of the mastery of substance and the completion of learning in the context of the study period. The completion of learning of learners is a component of the output obtained from the results of a learning process supported by parents, teachers and the environment. The success of online learning for the completion of learning, requires the cooperation of all components of Education itself.

scholarly journals Using Different Assessment Indicators in Supporting Online Learning

2021 ◽  
Author(s):  
Yew Kee Wong
Keyword(s):  
Online Learning ◽  
Statistical Analysis ◽  
Learning Outcomes ◽  
Learning Process ◽  
Learning Outcome ◽  
Learning System ◽  
Difficulty Level ◽  
Learning Methods ◽  
Analysis Algorithm ◽  
Efficient Learning

The assessment outcome for many online learning methods are based on the number of correct answers and than convert it into one final mark or grade. We discovered that when using online learning, we can extract more detail information from the learning process and these information are useful for the assessor to plan an effective and efficient learning model for the learner. Statistical analysis is an important part of an assessment when performing the online learning outcome. The assessment indicators include the difficulty level of the question, time spend in answering and the variation in choosing answer. In this paper we will present the findings of these assessment indicators and how it can improve the way the learner being assessed when using online learning system. We developed a statistical analysis algorithm which can assess the online learning outcomes more effectively using quantifiable measurements. A number of examples of using this statistical analysis algorithm are presented.

Distance and Online Learning

2017 ◽  
pp. 23-36
Author(s):  
Sema A. Kalaian
Keyword(s):  
Online Learning ◽  
Small Group ◽  
Online Courses ◽  
Group Learning ◽  
Small Group Learning ◽  
Learning Collaborative ◽  
Learning Methods ◽  
Course Content ◽  
And Mathematics ◽  
Teaching Learning

The aim of this chapter is to present a conceptual and practical overview of online learning pedagogies for the 21st century courses including science, technology, engineering, and mathematics (STEM) courses. Online learning and various alternative innovative forms of online small-group learning have been developed and implemented worldwide to replace or supplement the traditional face-to-face classroom instruction. Online teaching/learning using small-group learning methods such as problem-based learning, cooperative learning, collaborative learning methods, and team-based learning are examples of such innovative reform-based collaborative student-driven pedagogies that are covered in the chapter. These innovative 21st pedagogies make learning in online environments more stimulating, engaging, and motivating for students to deeply and meaningfully learn the course content and maximize their persistence in the web-based online courses.

Sign in / Sign up

Export Citation Format

Share Document