Direct Learning Method Research Articles

Adaptive stock trading strategies with deep reinforcement learning methods

The increasing complexity and dynamical property in stock markets are key challenges of the financial industry, in which inflexible trading strategies designed by experienced financial practitioners fail to achieve satisfactory performance in all market conditions. To meet this challenge, adaptive stock trading strategies with deep reinforcement learning methods are proposed. For the time-series nature of stock market data, the Gated Recurrent Unit (GRU) is applied to extract informative financial features, which can represent the intrinsic characteristics of the stock market for adaptive trading decisions. Furthermore, with the tailored design of state and action spaces, two trading strategies with reinforcement learning methods are proposed as GDQN (Gated Deep Q-learning trading strategy) and GDPG (Gated Deterministic Policy Gradient trading strategy). To verify the robustness and effectiveness of GDQN and GDPG, they are tested both in the trending and in the volatile stock market from different countries. Experimental results show that the proposed GDQN and GDPG not only outperform the Turtle trading strategy but also achieve more stable returns than a state-of-the-art direct reinforcement learning method, DRL trading strategy, in the volatile stock market. As far as the GDQN and the GDPG are compared, experimental results demonstrate that the GDPG with an actor-critic framework is more stable than the GDQN with a critic-only framework in the ever-evolving stock market.

Snowball Throwing: An English Learning Method to Teach Descriptive Writing

<span>Writing is a powerful multipurpose tool that helps us discover and explore more fully what we are thinking so that we learn as well as express our feelings and thoughts. One of reason that students lazy to written something because the methods are used by their teacher. Most of them still used direct learning method (extracurricular learning), the students were bored and they are did not enjoy when teaching learning process. So, that is why, in this study the writers try to use snowball throwing method to help students in learning descriptive writing. The data collection methods which used in this study are observation checklist and questionnaire checklist. Observation used to find out the implementation snowball throwing on students writing in descriptive text. The questionnaire was used to know the students’ responses after being taught descriptive text by using snowball throwing method. This study reveals that the highest aspect score of observation checklist in first observation is less, but in the second meeting, the writers found that highest aspect score of observation was good. Based on data questionnaire checklist, the writers found that the highest students’ responses were good. Based on this result data resource above, the writers concluded that the students could master writing after they were taught using snowball throwing in writing descriptive text. Finally, the writers suggested that the students had to practice their writing skill in many times. Besides, teacher had to have an interesting method in teaching writing descriptive text.</span>

Application of It-Based Learning Media to Improve the Basic Skills of Sewing Technology

The purpose of this study was to determine the activities of students in the basic learning activities of sewing technology using video media, to determine the learning outcomes of basic students of sewing technology using video media, and to determine students' responses to the basic learning activities of sewing technology using video media. This type of research is Classroom Action Research (CAR). The learning method is implemented using the direct learning method. Data collection techniques use observation to get data about student activities, performance tests to determine student learning outcomes, and questionnaires to determine student responses. Data analysis using quantitative descriptive. The results showed that student activities which include preliminary, core, and closing activities were said to be very good with an average value of 93% in the first cycle and rose to 96% in the second cycle. Student learning outcomes reached 73% grade completeness in the first cycle and rose to 82% in the second cycle. The average percentage of success in cycle I and II was 77% in the good category. The results of students' responses stated that they agreed that video media made them have a high willingness to participate in learning by 56%, with 47% of students strongly agree that learning with video media was very interesting and not boring, 58% agreed that it was easier to understand learning materials with video media, and 39% agree that video media motivate students to excel.

Peranan Statistika dan Pengembangan Karakter dalam Menghadapi Tantangan Era Revolusi Industri 4.0 dan Big Data pada SMAN 1 Praya

The development of science and technology provides changes to every aspect of human life including social, economic, educational and industrial changes which are now entering stage 4.0. The Era of the Industrial Revolution 4.0 is identical to the Internet of Things which produces Big Data that cannot be processed with conventional devices and requires special analysis. These changes require human resource development in science, education and character in order to continue to compete with the global world, especially the younger generation who will fill the industrial forward. The problem arises because most of the educational outcomes lack a link and match or a good match between tertiary education which causes students to feel wrong about their majors or the incompatibility of their needs and abilities in the industrial world which makes it difficult for them to find a work. Therefore, coaching efforts are needed so that students can be aware and prepare themselves to improve their quality both by increasing hardskills and soft skills to meet these needs. This community service activity is carried out by SMAN 1 Praya as one of the best high schools and is a reference school in West Nusa Tenggara. The method used is the direct learning method that is evaluated using self-assessment techniques conducted by students using google form. Evaluation results show an increase in students' knowledge of statistics and character development needed in the face of the industrial revolution 4.0 and Big Data after they have participated in this dedication activity.

A Direct Learning Approach for Neural Network Based Pre-Distortion for Coherent Nonlinear Optical Transmitter

High throughput coherent optical transmitters are key components in future optical communication infrastructure. However, these transmitters are often distorted with the nonlinearity of their components. A potential approach for compensating nonlinearity is by applying digital pre-distortion methods based on the Volterra series or one of its derivatives. However, the Volterra series-based solution is complex to implement, difficult to scale, and its simplified versions may not yield the desired performance. Recently digital pre distortion solutions based on neural networks were proposed, which may benefit from the generality of neural networks and can be more easily scaled. These solutions are often based on non-standard neural network architectures which require complex neurons-based architectures or being based on indirect training approach which suffer from noise enhancement. In this article, a novel method for neural network-based pre-distortion with direct learning is proposed. The direct learning with neural network does not assume a specific transmitter model and does not suffer from noise enhancement. The method assumes standard neural network inference architecture and is applied to a coherent nonlinear optical transmitted with long-short-term memory neural network. The overall performance and complexity of the direct learning method is compared with the indirect approach and with the Volterra series-based solution, showing significant advantage in performance, especially in cases of severe nonlinearity and noise conditions.

فاعلية بيئة تعلم إلکترونية شخصية قائمة على تطبيقات جوجل التفاعلية في تنمية مهارات التعامل مع شبکات الإنترنت لطلاب المرحلة الثانوية التجارية

فاعلية بيئة التعلم الإلکترونية الشخصية القائمة على تطبيقات جوجل التفاعلية في تنمية مهارات التعامل مع شبکات الإنترنت لطلاب المرحلة الثانوية التجارية. هدف هذا البحث إلى تنمية مهارات التعامل مع شبکات الإنترنت من خلال تصميم بيئة التعلم الإلکترونية الشخصية القائمة على تطبيقات جوجل التفاعلية في تنمية مهارات التعامل مع شبکات الإنترنت لطلاب المرحلة الثانوية التجارية، شملت عينة البحث (60) طالب وطالبة من طلاب الصف الثاني الثانوي التجاري من مدرسة الشهيد محمود الجيزي الثانوية التجارية المشترکة بإدارة المطرية التعليمية، وتم تقسيمهم بالتساوي إلي مجموعتين (تجريبية – ضابطة) تم اجراء الاختبار القبلي على کلا المجموعتين، ثم تخضع المجموعة التجريبية للدراسة باستخدام بيئة التعلم الإلکترونية الشخصية القائمة على تطبيقات جوجل التفاعلية (البيئة المصممة) ، بينما تلقت المجموعة الضابطة التدريس باستخدام طريقة التعلم التقليدي المباشر، ثم خضعت المجموعتين للاختبار البعدي لمقارنة نتائج المجموعتين، لقياس الأثر الذي أحدثه تطبيق المتغير المستقل، وقد ظهرت فروق بين المجموعتين في القياس البعدي لصالح المجموعة التجريبية، حيث ساهمت بيئة التعلم الإلکترونية الشخصية في تنمية مهارات التعامل مع شبکة الإنترنت لدى طلاب الصف الثاني للتعليم التجاري. الکلمات المفتاحية: بيئة التعلم الإلکترونية الشخصية – تطبيقات جوجل التفاعلية – شبکات الإنترنت. Title: The Effectiveness of Designing a Personalized E-Learning Environment Based on Google Interactive Applications in Developing the Skills of Dealing with the Internet for Business Secondary Students. Abstract: The aim of this study is to reveal the effectiveness of the design of the personal e-learning environment based on the interactive Google applications in developing the skills of dealing with the Internet for commercial secondary students, the sample of the study included (60) students of the second year secondary commercial students from Mahmoud Al-Gezi secondary commercial joint school in Mataria educational administration, They were divided equally into two groups one is an experimental and the other is a control group, applying pre test on both groups and then the experimental group is subjected to study using a personalized E-learning environment based on Google Interactive Applications (a designed environment). While the control group received instruction using the traditional direct learning method, the two groups were subjected to a post-test to compare the results of the two groups, to measure the impact of the application of the independent variable.There were differences between the two groups in the telemetry in favor of the experimental group. A personal e-learning environment contributedin developing the skills of dealing with the Internet for second grade students for business education. Keywords: Personal e-learning environment - Google interactive applications - Internet networks

Adaptive Basis Direct Learning Method for Predistortion of RF Power Amplifier

In this letter, an adaptive basis direct learning (DL) method was proposed for the linearization of power amplifiers (PAs). The proposed method can reduce the complexity and improve the stability performance of DL digital predistortion (DPD) structure. The experimental results show that the proposed method is more stable and can achieve improvement in normalized mean square error (NMSE) and adjacent channel power ratios (ACPRs) compared with the DL. In addition, the proposed method reduces the number of coefficients by 87%.

PENERAPAN MODEL CYCLE LEARNING 5E DAN METODE BRAINSTORMING TERHADAP KEMAMPUAN KOMUNIKASI MATEMATIS DITINJAU DARI MOTIVASI BELAJAR SISWA PADA MATERI SEGI EMPAT

<p>The purposes of this research were: (1) to understand does Cycle Learning 5E and Brainstorming Methods learning method produce the better mathematics communication capability rather than direct learning method, (2) to understand are students with high motivation have better mathematic communication skills than students with average motivation, are students with average motivation have better mathematic communication skills than students with low motivation, (3) to understand for students with high and average motivation does Cycle Learning 5E and Brainstorming method produce mathematic communication skills better than direct learning and for low motivation students does it produce the same mathematic skills with direct learning. (4) to compare on high with average motivation students, also both the average with low motivation students and high with low motivation students about how Cycle Learning 5E and Brainstorming methods gives them the better mathematic communication skills. This is a quantitative research with pseudo-experiment methods. The population of this research was all the VII grade students of Batang State 1 Junior High School. Sampling method was by Cluster Random Sampling technique. The data collection was using test method and questionnaire method. Data was analyzed by two ways Annova.</p>Result shows that: (1) Cycle Learning 5E and Brainstorming method gives better mathematic communication skills for students rather than direct learning; (2) students with high motivation have better mathematics communication skills than low motivation students, while high and average motivation students have the equal level of mathematic communication skills, and same goes to students with average and low motivation; (3) for all students with high, average and low motivation, the Cycle Learning 5E and Brainstorming method gives better mathematics communication skills than direct learning; (4) on teaching with Cycle Learning 5E and Brainstorming method, high motivation students have better mathematic communication skills rather than low motivation students, while they have equal communication skills with average motivation students. This condition happens also to the average and low motivation students.

Contextual teaching and learning of physics at elementary school

The student should be a master for science comprehension concepts, especially in material styles found in everyday life. The research aims to know influence approaching Contextual Teaching and Learning on the improve ability to comprehend the science concept. The experiment was carried out for 3 weeks involving 50 grade V elementary school students based on equal ability levels. pre-test and post-test refers to the falling motion material of an object due to gravitational force, how to enlarge and minimize friction force, as well as factors that influence the magnetic breakthrough force. As for each meeting an observation was carried out on 50 elementary school students conducted by the observer. The results of this test are processed and analyzed further using a t-test. The result showed that there is a significant difference between the ability of understanding concept students of experimental class using Contextual Teaching and Learning with control class students using direct learning method with pre-test post-test ≤ 0,05. Other than that the T test and N-gain of both classes showed the result of 0.000. That is approach Contextual Teaching and Learning influencing to improve understanding ability students’ concept science learning method in elementary school.

PERBANDINGAN METODE THINK-PAIR- SHARE DENGAN METODE CERAMAH TERHADAP HASIL BELAJAR SISWA PADA MATA PELAJARAN INSTALASI PERANGKAT JARINGAN LAN JURUSAN TKJ KELAS I SMK N 5 PADANG

The learning method in computer engineering and network at SMK N 5 Padang apply the direct learning method that have appreciation character and apply the speech method which is appropriate with the lesson,however, the student doesn’t have great motivation to follow the lesson and the lesson result is still under school grades standard.Because of that, it needs a learning model to make the student more active along with the increase of motivation and the student learning result in learning process. Based on the indication above, it conclude that the research question in this thesis is Is there any positive influence and significant toward the student learning result by using TPS method.The hypothesis of this research is there is the influence of TPS method toward the student learning result.This research uses the experiment phenomenological to see the influence of the student learning result by using TPS method and speech method.The population of the research is the student in first grade at SMKN 5 Padang 2013/2014 and the determining sample method is the purposive sampling with the result that the sample is the 1TKJ1A class by using speech method as the control class and 1TKJ1B by using TPS method as experiment group.Based on the research result, it seen that there is a comparison between the effect of stdent learning result experiment class and control class with 49,29% by using TPS method and 33,33% by using speech method.From the hypothesis test it is got that t count value is outside of Ho acceptance.It is mean that in the real situation the research shows that there is an influence of learning result and it is significant between TPS method and speech method in the lan network set installation lesson grade 1at SMK N 5 Padang. Keywords: Comparison, think pair share, speaking, learning result.

The effectiveness of the use of metacognition-based industrial electronic learning tools in growing higher order thinking skills (HOTS)

The aims of this research were to: (1) find out the response of students to metacognition-based industrial electronic learning in growing the ability of Higher Order Thinking Skills (HOTS) in vocational secondary schools, (2) describe the learning outcomes of the students in the field of industrial electronics based on metacognition with direct learning method, and (3) generate the effectiveness of metacognition-based industrial electronics learning in growing HOTS capabilities in vocational secondary schools. The research was Quasi-experimental type, while the research design used Pretest-Posttest Control Design with the research subject was the students of Public Vocational Secondary Schools in Makassar City in the field of Industrial Electronics Expertise. Data collection techniques were conducted by questionnaire and learning outcomes test. The result showed that: (1) the students’ response of experimental class was higher that control class, (2) the learning outcomes of experimental class was higher than control class, and (3) the use of metacognition-based tools for industrial electronics learning was effective. This learning method could grow the HOTS capabilities of the students in solving the problems of industrial electronics circuit.

Learning sparse structural changes in high-dimensional Markov networks

Recent years have seen an increasing popularity of learning the sparse changes in Markov Networks. Changes in the structure of Markov Networks reflect alternations of interactions between random variables under different regimes and provide insights into the underlying system. While each individual network structure can be complicated and difficult to learn, the overall change from one network to another can be simple. This intuition gave birth to an approach that directly learns the sparse changes without modelling and learning the individual (possibly dense) networks. In this paper, we review such a direct learning method with some latest developments along this line of research.

McMatMHKS: A direct multi-class matrixized learning machine

Multi-class classification learning can be implemented by the decomposition to binary classification or the direct techniques. The decomposition technique simplifies s the original learning problem into a set of binary subproblems, separately learns each one, and then combines their results to make a final decision. While the direct technique learns a set of multi-class classifiers by directly optimizing one single objective function. Plenty of empirical results have shown that the two techniques achieve comparable performance. However, both the techniques are mainly designed for vector-pattern samples at present. These traditional vector-pattern-oriented decomposition technique has been extended to a new type of matrix-pattern-oriented classifiers which obtain better learning performance and reduce the learning time–cost by utilizing the original structural information of the input matrix. To our best knowledge, no direct multi-class learning method for matrix pattern has been proposed so far. Therefore, this paper aims to propose a direct multi-class classification technique to compensate such a missing, which is a natural extension of the vector-based direct multi-class classification technique. Simultaneously, the left or right vector acting on matrix pattern in the multi-class matrixized objective function plays a role of a tradeoff parameter to balance the capacity of learning and generalization. Finally, based on the original binary-classifier Matrix-pattern-oriented Modified Ho-Kashyap classifier named MatMHKS, we design a corresponding Direct Multi-class Matrixized Learning Machine named McMatMHKS. It is the first direct multi-class classification technique for matrix patterns. To validate both feasibility and effectiveness of McMatMHKS, we conduct the comparative experiments on some benchmark datasets with two multi-class support vector machines and MatMHKS with the decomposition technique including both one-vs-one and one-vs-all. The results show that like its vector-oriented counterpart, McMatMHKS not only has comparable classification accuracy and AUC value, but also owns lower time complexity when compared with its corresponding decomposition machines.

Off-policy integral reinforcement learning optimal tracking control for continuous-time chaotic systems**Project supported by the National Natural Science Foundation of China (Grant Nos. 61304079 and 61374105), the Beijing Natural Science Foundation, China (Grant Nos. 4132078 and 4143065), the China Postdoctoral Science Foundation (Grant No. 2013M530527), the Fundamental Research Funds for the Central Universities, China (Grant No. FRF-TP-14-119A2), and the Open Research

This paper estimates an off-policy integral reinforcement learning (IRL) algorithm to obtain the optimal tracking control of unknown chaotic systems. Off-policy IRL can learn the solution of the HJB equation from the system data generated by an arbitrary control. Moreover, off-policy IRL can be regarded as a direct learning method, which avoids the identification of system dynamics. In this paper, the performance index function is first given based on the system tracking error and control error. For solving the Hamilton–Jacobi–Bellman (HJB) equation, an off-policy IRL algorithm is proposed. It is proven that the iterative control makes the tracking error system asymptotically stable, and the iterative performance index function is convergent. Simulation study demonstrates the effectiveness of the developed tracking control method.

Indirect Reinforcement Learning for Incident-responsive Ramp Control

A centralised strategy named indirect reinforcement learning ramp controller (IRLRC) has been developed in this paper to deal with ramp control problems for the congested traffic caused by incidents. IRLRC is developed on the basis of Dyna-Q architecture, under which a modified asymmetric cell transmission model (ACTM) and the standard Q-learning algorithm are combined together. The simulation-based test shows that compared with the no controlled situation, IRLRC can reduce the total travel time up to 24%, which outperforms the direct reinforcement learning (DRL) method with a reduction of 18% after the same number of iterations.

Adaptive Digital Predistortion for Wideband High Crest Factor Applications Based on the WACP Optimization Objective: An Extended Analysis

This paper provides an extended analysis of the adaptive digital predistortion technique initially proposed and conceptually overviewed in [1]. This digital predistortion technique is suitable for wideband high crest factor applications (digital audio broadcasting, digital video broadcasting-terrestrial & wideband code division multiple access high power transmitters) and overcomes the technical deficiencies of the traditional direct learning method. Specifically, predistortion filter parameter estimation is modeled as a generic mathematical optimization problem instead of a linear regression problem. In addition, the optimization objective is derived in the frequency instead of the time domain. A hybridly pruned Volterra series with memory is used to implement the predistortion filter. Hybrid pruning leads to a small optimization vector space, whilst the predistortion filter memory makes the method well suited to wideband applications. Given that predistortion filter parameter estimation does not rely on known test signals being injected into the transmitter, the method is on-air adaptive. Implementation aspects of the technique not covered in [1], but requiring extended coverage in this paper, include selection of mathematical optimization algorithms, predistortion filter memory estimation, weighted adjacent channel power weighting function application, and on-air adaption performance. Results obtained from actual hardware are presented.

NEURAL NETWORK-BASED LEARNING OF HUMAN SKILL IN HYDRAULIC ROBOTS

This paper proposes an alternative direct teaching and learning method of human skillful tasks for a hydraulic robot manipulator, which realizes the compatibility between the high power performance of hydraulic actuators and the fl exibility of the direct teaching system and is useful to replace the human skill with an appropriate robot controller. The proposed skill teaching process consists of two phases; the direct teaching phase and the off-line learning phase with a neural network applied to learn the skill. In the direct teaching phase, a human operator manipulates an impedancecontrolled robot directly and executes a task to obtain the training data. In the off-line learning phase, a neural network is trained to learn a control strategy included in human tasks. Then, the operator executes the task again in cooperation with a trained neural controller and learning processes are repeated alternately. After several learning processes a neural network is re-trained and the task performance of the controller is improved. The proposed approach is experimentally applied for a control problem of an inverted pendulum by using a 2-link hydraulic manipulator to show the validity of the system. The robustness of neural network controllers trained in different ways is also investigated.

The Interest Factor as Related to Methods of Introducing Beginners to Writing

Beginners are commonly introduced to writing in one two ways, either by the direct-learning method in which the child is presented with a model that he immediately attempts to copy, or the device method in which the child practices for a time with supplementary mechanical aids. The models for the first method may consist outlines simple objects which the child attempts to draw free-hand, followed by parts letters, whole letters, and words. Several forms mechanical devices which may be used are: furnishing the child with outlines familiar objects such as a coffee-pot and milk bottle and requiring him to fill in the out lines with colored pencils ; using sandpaper or emery outlines of letters or words which the child traces with his fingers; or having the beginner trace, with a pencil, grooved letters, or words. An other form is that in which the teacher grasps the child's hand and puts it through the writing movements, while still another form consists in giving the child a model copy over which he places a piece transparent paper and attempts to trace the letter or word through the paper. The use these and similar devices is based not only on the assumption that they have transfer value for the actual writing situation, but that they are interesting in them selves and relieve some the monotony and unpleasantness beginning handwriting. A recent experimental study concerning the effectiveness the direct-learning versus the device method teaching beginners to write, found that the direct-learning method produced by far the superior results.1 The present article, however, concerns itself with an analysis the interest factor these two methods observed in the experiment referred to. It must be mentioned that it was more difficult to analyze and state in quantitative terms the interest which the children manifested in the methods and devices they employed in learning to write than it was to measure the writing product itself. The experiment was planned primarily for the