A Study on the DRAM Market Using wavelet Analysis

スポンサーリンク

概要

• 論文の詳細を見る

• The characteristics of Dynamic Random Access Memory (DRAM) market movement are extracted by wavelet analysis. Four types of signals, the shipment and the spot price variation ratio of 4 MB and 16 MB DRAM are used as DRAM market data. The characteristics of the stages in the product life cycle of 4 MB and 16 MB DRAM are investigated using wavelet transformed market data. Furthermore, the aspect of wavelet transformation whereby wavelet coefficients still contain time information enables comparison among different signals. Therefore, the relationships between the trend changes of shipment and the spikes of the spot price variation ratio are captured.The discrete wavelet transformed plot represents the feature of respective resolutiojn level of wavelet coefficients.The absolute value of wavelet coefficients is depicted by the time-scale plot. The composition of the detail wavelet coefficients at each resolution level in a certain interval is captured by the energy, which is defined as the square of the detail wavelet coefficient. The shift in energy between resolution levels is depicted by an energy concentration ratio plot.

• 社団法人日本経営工学会の論文
• 2001-12-15

著者

• Suzuki Hideo

  University of Tsukuba

• Furuhata Masabumi

  Kpmg Consulting

• HURUHATA Masabumi

  KPMG Consulting

関連論文

• A ROBUST ENSEMBLE LEARNING USING ZERO-ONE LOSS FUNCTION
• A ROBUST BOOSTING METHOD FOR MISLABELED DATA
• A Study on the DRAM Market Using wavelet Analysis

スポンサーリンク