{
  "id": 294572,
  "title": "LSTM X Wavelets",
  "url": "/competitions/g-research-crypto-forecasting/discussion/294572",
  "author_name": "John Doe",
  "post_date": "2021-12-11T11:03:52.592000",
  "votes": 4,
  "comment_count": 0,
  "views": 0,
  "content": "<p>According to this article, WLSTM (LSTM with wavelets) outperforms LSTM and GRU.</p>\n<p><strong>Paper</strong><br>\n<a href=\"https://journals.plos.org/plosone/article/file?type=printable&amp;id=10.1371/journal.pone.0227222\" target=\"_blank\">Forecasting stock prices with long-short term memory neural network based on attention mechanism</a></p>\n<p><strong>Abstract</strong><br>\n<em>The stock market is known for its extreme complexity and volatility, and people are always\nlooking for an accurate and effective way to guide stock trading. Long short-term memory\n(LSTM) neural networks are developed by recurrent neural networks (RNN) and have significant application value in many fields. In addition, LSTM avoids long-term dependence\nissues due to its unique storage unit structure, and it helps predict financial time series.\nBased on LSTM and an attention mechanism, a wavelet transform is used to denoise historical stock data, extract and train its features, and establish the prediction model of a stock\nprice. We compared the results with the other three models, including the LSTM model, the\nLSTM model with wavelet denoising and the gated recurrent unit(GRU) neural network\nmodel on S&amp;P 500, DJIA, HSI datasets. Results from experiments on the S&amp;P 500 and\nDJIA datasets show that the coefficient of determination of the attention-based LSTM model\nis both higher than 0.94, and the mean square error of our model is both lower than 0.05.</em></p>",
  "messages": [
    {
      "id": 1614645,
      "postDate": "2021-12-11T11:03:52.593Z",
      "content": "<p>According to this article, WLSTM (LSTM with wavelets) outperforms LSTM and GRU.</p>\n<p><strong>Paper</strong><br>\n<a href=\"https://journals.plos.org/plosone/article/file?type=printable&amp;id=10.1371/journal.pone.0227222\" target=\"_blank\">Forecasting stock prices with long-short term memory neural network based on attention mechanism</a></p>\n<p><strong>Abstract</strong><br>\n<em>The stock market is known for its extreme complexity and volatility, and people are always\nlooking for an accurate and effective way to guide stock trading. Long short-term memory\n(LSTM) neural networks are developed by recurrent neural networks (RNN) and have significant application value in many fields. In addition, LSTM avoids long-term dependence\nissues due to its unique storage unit structure, and it helps predict financial time series.\nBased on LSTM and an attention mechanism, a wavelet transform is used to denoise historical stock data, extract and train its features, and establish the prediction model of a stock\nprice. We compared the results with the other three models, including the LSTM model, the\nLSTM model with wavelet denoising and the gated recurrent unit(GRU) neural network\nmodel on S&amp;P 500, DJIA, HSI datasets. Results from experiments on the S&amp;P 500 and\nDJIA datasets show that the coefficient of determination of the attention-based LSTM model\nis both higher than 0.94, and the mean square error of our model is both lower than 0.05.</em></p>",
      "rawMarkdown": "According to this article, WLSTM (LSTM with wavelets) outperforms LSTM and GRU.\n\n**Paper**\n[Forecasting stock prices with long-short term memory neural network based on attention mechanism](https://journals.plos.org/plosone/article/file?type=printable&id=10.1371/journal.pone.0227222)\n\n\n**Abstract**\n*The stock market is known for its extreme complexity and volatility, and people are always\nlooking for an accurate and effective way to guide stock trading. Long short-term memory\n(LSTM) neural networks are developed by recurrent neural networks (RNN) and have significant application value in many fields. In addition, LSTM avoids long-term dependence\nissues due to its unique storage unit structure, and it helps predict financial time series.\nBased on LSTM and an attention mechanism, a wavelet transform is used to denoise historical stock data, extract and train its features, and establish the prediction model of a stock\nprice. We compared the results with the other three models, including the LSTM model, the\nLSTM model with wavelet denoising and the gated recurrent unit(GRU) neural network\nmodel on S&P 500, DJIA, HSI datasets. Results from experiments on the S&P 500 and\nDJIA datasets show that the coefficient of determination of the attention-based LSTM model\nis both higher than 0.94, and the mean square error of our model is both lower than 0.05.*",
      "votes": 4
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1614645": "According to this article, WLSTM (LSTM with wavelets) outperforms LSTM and GRU.\n\n**Paper**\n[Forecasting stock prices with long-short term memory neural network based on attention mechanism](https://journals.plos.org/plosone/article/file?type=printable&id=10.1371/journal.pone.0227222)\n\n\n**Abstract**\n*The stock market is known for its extreme complexity and volatility, and people are always\nlooking for an accurate and effective way to guide stock trading. Long short-term memory\n(LSTM) neural networks are developed by recurrent neural networks (RNN) and have significant application value in many fields. In addition, LSTM avoids long-term dependence\nissues due to its unique storage unit structure, and it helps predict financial time series.\nBased on LSTM and an attention mechanism, a wavelet transform is used to denoise historical stock data, extract and train its features, and establish the prediction model of a stock\nprice. We compared the results with the other three models, including the LSTM model, the\nLSTM model with wavelet denoising and the gated recurrent unit(GRU) neural network\nmodel on S&P 500, DJIA, HSI datasets. Results from experiments on the S&P 500 and\nDJIA datasets show that the coefficient of determination of the attention-based LSTM model\nis both higher than 0.94, and the mean square error of our model is both lower than 0.05.*"
  }
}