Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model

碩士 === 輔仁大學 === 統計資訊學系應用統計碩士班 === 104 === This study applies Random Forest model to predict the price of Taiwan listed shares and its return rate. This study use the informations of institutional investors and index of main countries as daily data, and use technical indicators、futures index、exchan...

Full description

Bibliographic Details
Main Authors: LI,PEI-YING, 李佩螢
Other Authors: HUANG,HSIAO-YUN
Format: Others
Language:zh-TW
Published: 2016
Online Access:http://ndltd.ncl.edu.tw/handle/41448732017556051010
id ndltd-TW-104FJU00506008
record_format oai_dc
spelling ndltd-TW-104FJU005060082017-07-30T04:40:56Z http://ndltd.ncl.edu.tw/handle/41448732017556051010 Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model 以隨機森林時間序列模型為基礎之台灣上市股市即時交易策略之研究 LI,PEI-YING 李佩螢 碩士 輔仁大學 統計資訊學系應用統計碩士班 104 This study applies Random Forest model to predict the price of Taiwan listed shares and its return rate. This study use the informations of institutional investors and index of main countries as daily data, and use technical indicators、futures index、exchange rate、informations of Taiwan listed shares and index of Asia countries as intraday data. We integrate all variables through Random Forest method then generate a model and trading strategy. Final, let the study’s strategy and buy & hold strategy be compared. This study choose TSMC(2330)、LARGAN(3008)、ECLAT(1476) from Taiwan Top50 index to do backtesting. The empirical interval is from May 25, 2016 to June 3, 2016 with 8 trading days, 728 data. The empirical result shows that the model can fully learn with 1 or 2 days training data, and the input variable which highly influences the result is relationship between price and volume. Also, no matter what type of the stock price in empirical period is, the random forest model can predict the rising and falling section effectively. HUANG,HSIAO-YUN 黃孝雲 2016 學位論文 ; thesis 57 zh-TW
collection NDLTD
language zh-TW
format Others
sources NDLTD
description 碩士 === 輔仁大學 === 統計資訊學系應用統計碩士班 === 104 === This study applies Random Forest model to predict the price of Taiwan listed shares and its return rate. This study use the informations of institutional investors and index of main countries as daily data, and use technical indicators、futures index、exchange rate、informations of Taiwan listed shares and index of Asia countries as intraday data. We integrate all variables through Random Forest method then generate a model and trading strategy. Final, let the study’s strategy and buy & hold strategy be compared. This study choose TSMC(2330)、LARGAN(3008)、ECLAT(1476) from Taiwan Top50 index to do backtesting. The empirical interval is from May 25, 2016 to June 3, 2016 with 8 trading days, 728 data. The empirical result shows that the model can fully learn with 1 or 2 days training data, and the input variable which highly influences the result is relationship between price and volume. Also, no matter what type of the stock price in empirical period is, the random forest model can predict the rising and falling section effectively.
author2 HUANG,HSIAO-YUN
author_facet HUANG,HSIAO-YUN
LI,PEI-YING
李佩螢
author LI,PEI-YING
李佩螢
spellingShingle LI,PEI-YING
李佩螢
Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model
author_sort LI,PEI-YING
title Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model
title_short Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model
title_full Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model
title_fullStr Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model
title_full_unstemmed Taiwan Stock Market Intraday Trading Strategy Using Random Forest Time Series Model
title_sort taiwan stock market intraday trading strategy using random forest time series model
publishDate 2016
url http://ndltd.ncl.edu.tw/handle/41448732017556051010
work_keys_str_mv AT lipeiying taiwanstockmarketintradaytradingstrategyusingrandomforesttimeseriesmodel
AT lǐpèiyíng taiwanstockmarketintradaytradingstrategyusingrandomforesttimeseriesmodel
AT lipeiying yǐsuíjīsēnlínshíjiānxùlièmóxíngwèijīchǔzhītáiwānshàngshìgǔshìjíshíjiāoyìcèlüèzhīyánjiū
AT lǐpèiyíng yǐsuíjīsēnlínshíjiānxùlièmóxíngwèijīchǔzhītáiwānshàngshìgǔshìjíshíjiāoyìcèlüèzhīyánjiū
_version_ 1718508094591336448