• 제목/요약/키워드: Approaches to Learning

검색결과 999건 처리시간 0.027초

Modeling and Stimulating Node Cooperation in Wireless Ad Hoc Networks

  • Arghavani, Abbas;Arghavani, Mahdi;Sargazi, Abolfazl;Ahmadi, Mahmood
    • ETRI Journal
    • /
    • 제37권1호
    • /
    • pp.77-87
    • /
    • 2015
  • In wireless networks, cooperation is necessary for many protocols, such as routing, clock synchronization, and security. It is known that cooperator nodes suffer greatly from problems such as increasing energy consumption. Therefore, rational nodes have no incentive to cooperatively forward traffic for others. A rational node is different from a malicious node. It is a node that makes the best decision in each state (cooperate or non-cooperate). In this paper, game theory is used to analyze the cooperation between nodes. An evolutionary game has been investigated using two nodes, and their strategies have been compared to find the best one. Subsequently, two approaches, one based on a genetic algorithm (GA) and the other on learning automata (LA), are presented to incite nodes for cooperating in a noisy environment. As you will see later, the GA strategy is able to disable the effect of noise by using a big enough chromosome; however, it cannot persuade nodes to cooperate in a noisefree environment. Unlike the GA strategy, the LA strategy shows good results in a noise-free environment because it has good agreement in cooperation-based strategies in both types of environment (noise-free and noisy).

Combining Dynamic Time Warping and Single Hidden Layer Feedforward Neural Networks for Temporal Sign Language Recognition

  • Thi, Ngoc Anh Nguyen;Yang, Hyung-Jeong;Kim, Sun-Hee;Kim, Soo-Hyung
    • International Journal of Contents
    • /
    • 제7권1호
    • /
    • pp.14-22
    • /
    • 2011
  • Temporal Sign Language Recognition (TSLR) from hand motion is an active area of gesture recognition research in facilitating efficient communication with deaf people. TSLR systems consist of two stages: a motion sensing step which extracts useful features from signers' motion and a classification process which classifies these features as a performed sign. This work focuses on two of the research problems, namely unknown time varying signal of sign languages in feature extraction stage and computing complexity and time consumption in classification stage due to a very large sign sequences database. In this paper, we propose a combination of Dynamic Time Warping (DTW) and application of the Single hidden Layer Feedforward Neural networks (SLFNs) trained by Extreme Learning Machine (ELM) to cope the limitations. DTW has several advantages over other approaches in that it can align the length of the time series data to a same prior size, while ELM is a useful technique for classifying these warped features. Our experiment demonstrates the efficiency of the proposed method with the recognition accuracy up to 98.67%. The proposed approach can be generalized to more detailed measurements so as to recognize hand gestures, body motion and facial expression.

딥 러닝과 마르코프 랜덤필드를 이용한 동영상 내 그림자 검출 (Moving Shadow Detection using Deep Learning and Markov Random Field)

  • 이종택;강현우;임길택
    • 한국멀티미디어학회논문지
    • /
    • 제18권12호
    • /
    • pp.1432-1438
    • /
    • 2015
  • We present a methodology to detect moving shadows in video sequences, which is considered as a challenging and critical problem in the most visual surveillance systems since 1980s. While most previous moving shadow detection methods used hand-crafted features such as chromaticity, physical properties, geometry, or combination thereof, our method can automatically learn features to classify whether image segments are shadow or foreground by using a deep learning architecture. Furthermore, applying Markov Random Field enables our system to refine our shadow detection results to improve its performance. Our algorithm is applied to five different challenging datasets of moving shadow detection, and its performance is comparable to that of state-of-the-art approaches.

전문가시스템과 신경회로망에 의한 축사환경개선시스템 (Troubleshooting System for Environmental Problems in a Livestock Building Using an Expert System and a Neural Network)

  • 손정익;;김문기
    • 한국농공학회지
    • /
    • 제36권1호
    • /
    • pp.95-102
    • /
    • 1994
  • Since parameters influencing the indoor environment of livestock building interrelate so complicatedly, it is of great difficulty to identify the exact cause of environmental problems in a livestock building. Therefore, the approaches for the problem solving based on experience not numerical calculation will be helpful to the management of livestock building This study was attempt to develop the decision supporting system to diagnose environmen- tal problems in a livestock building based on an expert system and a neural network. HClips$^3$), attaching the Hangeul user interface to Clips which is known as a powerful shell for develop- ing expert system, was used. The multilayer perceptron consisting of 4 layers including back propagation learning algorithm was adpoted, which was rapidly converged within the allowable range at 50,000 learning sweeps. The expert system and neural network seemed to work well for this specific application, providing proper suggestions for some environmental problems: particularly, the neural net- work trained by an environmental problem and its corresponding answer with certainty factor, produced the same results as those by expert system.

  • PDF

A Synaptic Model for Pain: Long-Term Potentiation in the Anterior Cingulate Cortex

  • Zhuo, Min
    • Molecules and Cells
    • /
    • 제23권3호
    • /
    • pp.259-271
    • /
    • 2007
  • Investigation of molecular and cellular mechanisms of synaptic plasticity is the major focus of many neuroscientists. There are two major reasons for searching new genes and molecules contributing to central plasticity: first, it provides basic neural mechanism for learning and memory, a key function of the brain; second, it provides new targets for treating brain-related disease. Long-term potentiation (LTP), mostly intensely studies in the hippocampus and amygdala, is proposed to be a cellular model for learning and memory. Although it remains difficult to understand the roles of LTP in hippocampus-related memory, a role of LTP in fear, a simplified form of memory, has been established. Here, I will review recent cellular studies of LTP in the anterior cingulate cortex (ACC) and then compare studies in vivo and in vitro LTP by genetic/pharmacological approaches. I propose that ACC LTP may serve as a cellular model for studying central sensitization that related to chronic pain, as well as pain-related cognitive emotional disorders. Understanding signaling pathways related to ACC LTP may help us to identify novel drug target for various mental disorders.

Effects of Action Observation Training Combied with Auditory Cueing on Gait Ability in Patients with Stroke: a Preliminary Pilot Study

  • Kim, Hyeong-Min;Son, Sung-Min;Ko, Yu-Min
    • The Journal of Korean Physical Therapy
    • /
    • 제34권3호
    • /
    • pp.98-103
    • /
    • 2022
  • Purpose: New therapeutic approaches have emerged to improve gait ability in patients with brain damage, such as action observation learning (AOT), auditory cueing, motor imagery etc. We attempted to investigate the effects of AOT with auditory cueing (AOTAC) on gait function in patients with stroke. Methods: The eighteen stroke patients with a unilateral hemiparesis were randomly divided into three groups; the AOTAC, AOT, and control groups. The AOTAC group (n=8) received training via observing a video that showed normal gait with sound of footsteps as an auditory cue; the AOT group (n=6) receive action observation without auditory stimulation; the control group (n=5) observed the landscape video image. Intervention time of three groups was 30 minutes per day, five times a week, for four weeks. Gait parameters, such as cadence, velocity, stride length, stance phase, and swing phase were collected in all patients before and after each training session. Results: Significant differences were observed among the three groups with respect to the parameters, such as cadence, velocity, stride length, and stance/swing phase. Post-hoc analysis indicated that the AOTAC group had a greater significant change in all of parameters, compared with the AOT and control groups. Conclusion: Our findings suggest that AOTAC may be an effective therapeutic approach to improve gait symmetry and function in patients with stroke. We believe that this effect is attributable to the change of cortical excitability on motor related to cortical areas.

딥러닝 기반 BIM 부재 자동분류 학습모델의 성능 향상을 위한 Ensemble 모델 구축에 관한 연구 (Advanced Approach for Performance Improvement of Deep Learningbased BIM Elements Classification Model Using Ensemble Model)

  • 김시현;이원복;유영수;구본상
    • 한국BIM학회 논문집
    • /
    • 제12권2호
    • /
    • pp.12-25
    • /
    • 2022
  • To increase the usability of Building Information Modeling (BIM) in construction projects, it is critical to ensure the interoperability of data between heterogeneous BIM software. The Industry Foundation Classes (IFC), an international ISO format, has been established for this purpose, but due to its structural complexity, geometric information and properties are not always transmitted correctly. Recently, deep learning approaches have been used to learn the shapes of the BIM elements and thereby verify the mapping between BIM elements and IFC entities. These models performed well for elements with distinct shapes but were limited when their shapes were highly similar. This study proposed a method to improve the performance of the element type classification by using an Ensemble model that leverages not only shapes characteristics but also the relational information between individual BIM elements. The accuracy of the Ensemble model, which merges MVCNN and MLP, was improved 0.03 compared to the existing deep learning model that only learned shape information.

A Stay Detection Algorithm Using GPS Trajectory and Points of Interest Data

  • Eunchong Koh;Changhoon Lyu;Goya Choi;Kye-Dong Jung;Soonchul Kwon;Chigon Hwang
    • International Journal of Internet, Broadcasting and Communication
    • /
    • 제15권3호
    • /
    • pp.176-184
    • /
    • 2023
  • Points of interest (POIs) are widely used in tourism recommendations and to provide information about areas of interest. Currently, situation judgement using POI and GPS data is mainly rule-based. However, this approach has the limitation that inferences can only be made using predefined POI information. In this study, we propose an algorithm that uses POI data, GPS data, and schedule information to calculate the current speed, location, schedule matching, movement trajectory, and POI coverage, and uses machine learning to determine whether to stay or go. Based on the input data, the clustered information is labelled by k-means algorithm as unsupervised learning. This result is trained as the input vector of the SVM model to calculate the probability of moving and staying. Therefore, in this study, we implemented an algorithm that can adjust the schedule using the travel schedule, POI data, and GPS information. The results show that the algorithm does not rely on predefined information, but can make judgements using GPS data and POI data in real time, which is more flexible and reliable than traditional rule-based approaches. Therefore, this study can optimize tourism scheduling. Therefore, the stay detection algorithm using GPS movement trajectories and POIs developed in this study provides important information for tourism schedule planning and is expected to provide much value for tourism services.

Generative Interactive Psychotherapy Expert (GIPE) Bot

  • Ayesheh Ahrari Khalaf;Aisha Hassan Abdalla Hashim;Akeem Olowolayemo;Rashidah Funke Olanrewaju
    • International Journal of Computer Science & Network Security
    • /
    • 제23권4호
    • /
    • pp.15-24
    • /
    • 2023
  • One of the objectives and aspirations of scientists and engineers ever since the development of computers has been to interact naturally with machines. Hence features of artificial intelligence (AI) like natural language processing and natural language generation were developed. The field of AI that is thought to be expanding the fastest is interactive conversational systems. Numerous businesses have created various Virtual Personal Assistants (VPAs) using these technologies, including Apple's Siri, Amazon's Alexa, and Google Assistant, among others. Even though many chatbots have been introduced through the years to diagnose or treat psychological disorders, we are yet to have a user-friendly chatbot available. A smart generative cognitive behavioral therapy with spoken dialogue systems support was then developed using a model Persona Perception (P2) bot with Generative Pre-trained Transformer-2 (GPT-2). The model was then implemented using modern technologies in VPAs like voice recognition, Natural Language Understanding (NLU), and text-to-speech. This system is a magnificent device to help with voice-based systems because it can have therapeutic discussions with the users utilizing text and vocal interactive user experience.

딥러닝 기반 뉴로사이언스 마이닝 기법을 이용한 고객 매력/유용성 인지 (CAUP) 예측 성능에 관한 탐색적 연구: Dark vs Light 사용자 인터페이스 (UI)를 중심으로 (Exploring the Performance of Deep Learning-Driven Neuroscience Mining in Predicting CAUP (Consumer's Attractiveness/Usefulness Perception): Emphasis on Dark vs Light UI Modes)

  • 김민경;;이건창
    • 한국컴퓨터정보학회:학술대회논문집
    • /
    • 한국컴퓨터정보학회 2022년도 제66차 하계학술대회논문집 30권2호
    • /
    • pp.19-22
    • /
    • 2022
  • In this work, we studied consumers' attractiveness/usefulness perceptions (CAUP) of online commerce product photos when exposed to alternative dark/light user interface (UI) modes. We analyzed time-series EEG data from 31 individuals and performed neuroscience mining (NSM) to ascertain (a) how the CAUP of products differs among UI modes; and (b) which deep learning model provides the most accurate assessment of such neuroscience mining (NSM) business difficulties. The dark UI style increased the CAUP of the products displayed and was predicted with the greatest accuracy using a unique EEG power spectra separated wave brainwave 2D-ConvLSTM model. Then, using relative importance analysis, we used this model to determine the most relevant power spectra. Our findings are considered to contribute to the discovery of objective truths about online customers' reactions to various user interface modes used by various online marketplaces that cannot be uncovered through more traditional research approaches like as surveys.

  • PDF