• Title/Summary/Keyword: feature models

Search Result 1,135, Processing Time 0.022 seconds

A dynamic procedure for defection detection and prevention based on SOM and a Markov chain

  • Kim, Young-ae;Song, Hee-seok;Kim, Soung-hie
    • Proceedings of the KAIS Fall Conference
    • /
    • 2003.11a
    • /
    • pp.141-148
    • /
    • 2003
  • Customer retention is a common concern for many industries and a critical issue for the survival in today's greatly compressed marketplace. Current customer retention models only focus on detection of potential defectors based on the likelihood of defection by using demographic and customer profile information. In this paper, we propose a dynamic procedure for defection detection and prevention using past and current customer behavior by utilizing SOM and Markov chain. The basic idea originates from the observation that a customer has a tendency to change his behavior (i.e. trim-out his usage volumes) before his eventual withdrawal. This gradual pulling out process offers the company the opportunity to detect the defection signals. With this approach, we have two significant benefits compared with existing defection detection studies. First, our procedure can predict when the potential defectors could withdraw and this feature helps to give marketing managers ample lead-time for preparing defection prevention plans. The second benefit is that our approach can provide a procedure for not only defection detection but also defection prevention, which could suggest the desirable behavior state for the next period so as to lower the likelihood of defection. We applied our dynamic procedure for defection detection and prevention to the online gaming industry. Our suggested procedure could predict potential defectors without deterioration of prediction accuracy compared to that of the MLP neural network and DT.

  • PDF

Hybrid Fuzzy Least Squares Support Vector Machine Regression for Crisp Input and Fuzzy Output

  • Shim, Joo-Yong;Seok, Kyung-Ha;Hwang, Chang-Ha
    • Communications for Statistical Applications and Methods
    • /
    • v.17 no.2
    • /
    • pp.141-151
    • /
    • 2010
  • Hybrid fuzzy regression analysis is used for integrating randomness and fuzziness into a regression model. Least squares support vector machine(LS-SVM) has been very successful in pattern recognition and function estimation problems for crisp data. This paper proposes a new method to evaluate hybrid fuzzy linear and nonlinear regression models with crisp inputs and fuzzy output using weighted fuzzy arithmetic(WFA) and LS-SVM. LS-SVM allows us to perform fuzzy nonlinear regression analysis by constructing a fuzzy linear regression function in a high dimensional feature space. The proposed method is not computationally expensive since its solution is obtained from a simple linear equation system. In particular, this method is a very attractive approach to modeling nonlinear data, and is nonparametric method in the sense that we do not have to assume the underlying model function for fuzzy nonlinear regression model with crisp inputs and fuzzy output. Experimental results are then presented which indicate the performance of this method.

Performance Improvement of Automatic Speech Segmentation and Labeling System (자동 음성분할 및 레이블링 시스템의 성능향상)

  • Hong Seong Tae;Kim Je-U;Kim Hyeong-Sun
    • MALSORI
    • /
    • no.35_36
    • /
    • pp.175-188
    • /
    • 1998
  • Database segmented and labeled up to phoneme level plays an important role in phonetic research and speech engineering. However, it usually requires manual segmentation and labeling, which is time-consuming and may also lead to inconsistent consequences. Automatic segmentation and labeling can be introduced to solve these problems. In this paper, we investigate a method to improve the performance of automatic segmentation and labeling system, where Spectral Variation Function(SVF), modification of silence model, and use of energy variations in postprocessing stage are considered. In this paper, SVF is applied in three ways: (1) addition to feature parameters, (2) postprocessing of phoneme boundaries, (3) restricting the Viterbi path so that the resulting phoneme boundaries may be located in frames around SVF peaks. In the postprocessing stage, positions with greatest energy variation during transitional period between silence and other phonemes were used to modify boundaries. In order to evaluate the performance of the system, we used 452 phonetically balanced word(PBW) database for training phoneme models and phonetically balanced sentence(PBS) database for testing. According to our experiments, 83.1% (6.2% improved) and 95.8% (0.9% improved) of phoneme boundaries were within 20ms and 40ms of the manually segmented boundaries, respectively.

  • PDF

Modeling of the Elasto-plastic Deformation Behavior of Two-Dimensional Anisotropic Foam under Compressive Loads using Voronoi Cells (보로노이 셀을 이용한 2 원 비등방성 폼 구조 모델링 및 탄소성 압축변형 해석)

  • Han, Won-Hee;Choi, Byoung-Ho;Kim, Il-Hyun;Lee, Jeong-Moo
    • Journal of the Korean Society for Precision Engineering
    • /
    • v.29 no.7
    • /
    • pp.785-792
    • /
    • 2012
  • Foam structure is usually hard to model due to the complexity of the geometry of cells. So, many simplified models to represent complicated foam structures have been proposed, but most of them are not actually describe the random feature of the cell structure well. So, in this study, two dimensional isotropic and anisotropic closed cell structures of the foam were modeled using the concept of Voronoi cells. The elasto-plastic deformation behavior under compressive loads was investigated by finitie element analysis, and the results were compared with ideal honeycomb structure. In addition, the effect of anisotropy of Voronoi cell structures of the foam on Young's modulus and yield stress under compressive loads was studied.

Deconvolution Pixel Layer Based Semantic Segmentation for Street View Images (디컨볼루션 픽셀층 기반의 도로 이미지의 의미론적 분할)

  • Wahid, Abdul;Lee, Hyo Jong
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • 2019.05a
    • /
    • pp.515-518
    • /
    • 2019
  • Semantic segmentation has remained as a challenging problem in the field of computer vision. Given the immense power of Convolution Neural Network (CNN) models, many complex problems have been solved in computer vision. Semantic segmentation is the challenge of classifying several pixels of an image into one category. With the help of convolution neural networks, we have witnessed prolific results over the time. We propose a convolutional neural network model which uses Fully CNN with deconvolutional pixel layers. The goal is to create a hierarchy of features while the fully convolutional model does the primary learning and later deconvolutional model visually segments the target image. The proposed approach creates a direct link among the several adjacent pixels in the resulting feature maps. It also preserves the spatial features such as corners and edges in images and hence adding more accuracy to the resulting outputs. We test our algorithm on Karlsruhe Institute of Technology and Toyota Technologies Institute (KITTI) street view data set. Our method achieves an mIoU accuracy of 92.04 %.

Buckling response with stretching effect of carbon nanotube-reinforced composite beams resting on elastic foundation

  • Khelifa, Zoubida;Hadji, Lazreg;Daouadji, Tahar Hassaine;Bourada, Mohamed
    • Structural Engineering and Mechanics
    • /
    • v.67 no.2
    • /
    • pp.125-130
    • /
    • 2018
  • This study deals with buckling analysis with stretching effect of functionally graded carbon nanotube-reinforced composite beams resting on an elastic foundation. The single-walled carbon nanotubes (SWCNTs) are aligned and distributed in polymeric matrix with different patterns of reinforcement. The material properties of the CNTRC beams are estimated by using the rule of mixture. The significant feature of this model is that, in addition to including the shear deformation effect and stretching effect it deals with only 4 unknowns without including a shear correction factor. The equilibrium equations have been obtained using the principle of virtual displacements. The mathematical models provided in this paper are numerically validated by comparison with some available results. New results of buckling analyses of CNTRC beams based on the present theory with stretching effect is presented and discussed in details. the effects of different parameters of the beam on the buckling responses of CNTRC beam are discussed.

Using SG Arrays for Hydrology in Comparison with GRACE Satellite Data, with Extension to Seismic and Volcanic Hazards

  • Crossley David;Hinderer Jacques
    • Korean Journal of Remote Sensing
    • /
    • v.21 no.1
    • /
    • pp.31-49
    • /
    • 2005
  • We first review some history of the Global Geodynamics Project (GGP), particularly in the progress of ground-satellite gravity comparisons. The GGP Satellite Project has involved the measurement of ground-based superconducting gravimeters (SGs) in Europe for several years and we make quantitative comparisons with the latest satellite GRACE data and hydrological models. The primary goal is to recover information about seasonal hydrology cycles, and we find a good correlation at the microgal level between the data and modeling. One interesting feature of the data is low soil moisture resulting from the European heat wave in 2003. An issue with the ground-based stations is the possibility of mass variations in the soil above a station, and particularly for underground stations these have to be modeled precisely. Based on this work with a regional array, we estimate the effectiveness of future SG arrays to measure co-seismic deformation and silent-slip events. Finally we consider gravity surveys in volcanic areas, and predict the accuracy in modeling subsurface density variations over time periods from months to years.

Two Modified Z-Source Inverter Topologies - Solutions to Start-Up Dc-Link Voltage Overshoot and Source Current Ripple

  • Bharatkumar, Dave Heema;Singh, Dheerendra;Bansal, Hari Om
    • Journal of Power Electronics
    • /
    • v.19 no.6
    • /
    • pp.1351-1365
    • /
    • 2019
  • This paper proposes two modified Z-source inverter topologies, namely an embedded L-Z-source inverter (EL-ZSI) and a coupled inductor L-Z source inverter (CL-ZSI). The proposed topologies offer a high voltage gain with a reduced passive component count and reduction in source current ripple when compared to conventional ZSI topologies. Additionally, they prevent overshoot in the dc-link voltage by suppressing heavy inrush currents. This feature reduces the transition time to reach the peak value of the dc-link voltage, and reduces the risk of component failure and overrating due to the inrush current. EL-ZSI and CL-ZSI possess all of the inherent advantages of the conventional L-ZSI topology while eliminating its drawbacks. To verify the effectiveness of the proposed topologies, MATLAB/Simulink models and scaled down laboratory prototypes were constructed. Experiments were performed at a low shoot through duty ratio of 0.1 and a modulation index as high as 0.9 to obtain a peak dc-link voltage of 53 V. This paper demonstrates the superiority of the proposed topologies over conventional ZSI topologies through a detailed comparative analysis. Moreover, experimental results verify that the proposed topologies would be advantageous for renewable energy source applications since they provide voltage gain enhancement, inrush current, dc-link voltage overshoot suppression and a reduction of the peak to peak source current ripple.

Sparse Matrix Compression Technique and Hardware Design for Lightweight Deep Learning Accelerators (경량 딥러닝 가속기를 위한 희소 행렬 압축 기법 및 하드웨어 설계)

  • Kim, Sunhee;Shin, Dongyeob;Lim, Yong-Seok
    • Journal of Korea Society of Digital Industry and Information Management
    • /
    • v.17 no.4
    • /
    • pp.53-62
    • /
    • 2021
  • Deep learning models such as convolutional neural networks and recurrent neual networks process a huge amounts of data, so they require a lot of storage and consume a lot of time and power due to memory access. Recently, research is being conducted to reduce memory usage and access by compressing data using the feature that many of deep learning data are highly sparse and localized. In this paper, we propose a compression-decompression method of storing only the non-zero data and the location information of the non-zero data excluding zero data. In order to make the location information of non-zero data, the matrix data is divided into sections uniformly. And whether there is non-zero data in the corresponding section is indicated. In this case, section division is not executed only once, but repeatedly executed, and location information is stored in each step. Therefore, it can be properly compressed according to the ratio and distribution of zero data. In addition, we propose a hardware structure that enables compression and decompression without complex operations. It was designed and verified with Verilog, and it was confirmed that it can be used in hardware deep learning accelerators.

Two-Stream Convolutional Neural Network for Video Action Recognition

  • Qiao, Han;Liu, Shuang;Xu, Qingzhen;Liu, Shouqiang;Yang, Wanggan
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.15 no.10
    • /
    • pp.3668-3684
    • /
    • 2021
  • Video action recognition is widely used in video surveillance, behavior detection, human-computer interaction, medically assisted diagnosis and motion analysis. However, video action recognition can be disturbed by many factors, such as background, illumination and so on. Two-stream convolutional neural network uses the video spatial and temporal models to train separately, and performs fusion at the output end. The multi segment Two-Stream convolutional neural network model trains temporal and spatial information from the video to extract their feature and fuse them, then determine the category of video action. Google Xception model and the transfer learning is adopted in this paper, and the Xception model which trained on ImageNet is used as the initial weight. It greatly overcomes the problem of model underfitting caused by insufficient video behavior dataset, and it can effectively reduce the influence of various factors in the video. This way also greatly improves the accuracy and reduces the training time. What's more, to make up for the shortage of dataset, the kinetics400 dataset was used for pre-training, which greatly improved the accuracy of the model. In this applied research, through continuous efforts, the expected goal is basically achieved, and according to the study and research, the design of the original dual-flow model is improved.