WEKO3
アイテム
対象の値動きの方向性に着目した経済時系列予測へのサポートベクターマシンの応用に関する研究
http://hdl.handle.net/2261/28804
http://hdl.handle.net/2261/288045d3b73db-8a42-4010-ad8f-872840ebdb26
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
48066432.pdf (1.2 MB)
|
|
| Item type | 学位論文 / Thesis or Dissertation(1) | |||||
|---|---|---|---|---|---|---|
| 公開日 | 2011-08-08 | |||||
| タイトル | ||||||
| タイトル | 対象の値動きの方向性に着目した経済時系列予測へのサポートベクターマシンの応用に関する研究 | |||||
| 言語 | ||||||
| 言語 | jpn | |||||
| キーワード | ||||||
| 主題Scheme | Other | |||||
| 主題 | Support Vector Machine | |||||
| 資源タイプ | ||||||
| 資源 | http://purl.org/coar/resource_type/c_46ec | |||||
| タイプ | thesis | |||||
| その他のタイトル | ||||||
| その他のタイトル | Research regarding the Application of Support Vector Machines to Predict the Direction of Price Changes in Economic Time Series | |||||
| 著者 |
中田, 貴之
× 中田, 貴之 |
|||||
| 著者別名 | ||||||
| 識別子Scheme | WEKO | |||||
| 識別子 | 5721 | |||||
| 姓名 | Nakata, Takayuki | |||||
| 著者所属 | ||||||
| 著者所属 | 大学院情報理工学系研究科電子情報学専攻 | |||||
| 著者所属 | ||||||
| 著者所属 | Graduate School of Information Science and Technology Department of Information and Communication Engineering The University of Tokyo | |||||
| Abstract | ||||||
| 内容記述タイプ | Abstract | |||||
| 内容記述 | Today, time series data are predicted using various methods. The main technique currently used to identify a time series has been adjusting the coefficient parameter in order to minimize the disparity between predicted data values. This is a logical method from the viewpoint of time series data prediction, but it does not suit the needs of economic time series prediction. With economic time series, like stock index data, the final goal of prediction is to invest more efficiently. To accomplish efficient investment, it is more important to raise the hit ratio in predicting the direction of price changes rather than to minimize the incongruity in data trends. This research uses Support Vector Machines (SVM), a method of pattern recognition, and the direction-of-price-change hit ratio, as an index for evaluating economic time series data prediction. The empirical results detailed in this thesis reveal that short term or midterm predictions by SVM are useful for index investments. This method can be generalized for investing in many different financial institutions. | |||||
| 書誌情報 | 発行日 2008-03 | |||||
| 日本十進分類法 | ||||||
| 主題Scheme | NDC | |||||
| 主題 | 548 | |||||
| 学位名 | ||||||
| 学位名 | 修士(情報理工学) | |||||
| 学位 | ||||||
| 値 | master | |||||
| 研究科・専攻 | ||||||
| 情報理工学系研究科電子情報学専攻 | ||||||
| 学位授与年月日 | ||||||
| 学位授与年月日 | 2008-03-24 | |||||
