Skip to content

Nasdaq Time Series Forecasting using PatchTSM Transformers model (Hugging Face)

License

Notifications You must be signed in to change notification settings

prathyyyyy/nasdaq-stock-prediction

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

nasdaq-stock-prediction Using PatchTSM Transformer Model (Hugging Face)

1. Introduction:

Nasdaq is one of the most prominent stock exchanges globally, recognized for its comprehensive electronic trading platform and a significant concentration of technology and biotech companies. Established on February 8, 1971, and headquartered in New York City, the Nasdaq operates entirely electronically, facilitating fast, efficient, and transparent trading. It boasts a substantial market capitalization, often rivaling that of the New York Stock Exchange (NYSE), and lists many of the world's leading tech giants, such as Apple, Microsoft, Amazon, and Google.

The Nasdaq's trading hours extend beyond the regular session of 9:30 AM to 4:00 PM Eastern Time (ET) to include pre-market and after-hours trading, providing investors with extended opportunities to engage in the market. This flexibility, coupled with its electronic trading capabilities, makes Nasdaq a vital hub for innovation and a critical player in the global financial markets. The performance of its indices, particularly the Nasdaq-100, is widely regarded as a barometer of the health of the technology sector and overall investor sentiment towards growth stocks.

This dataset has 9 million rows

2. Framework And Tools

Framework - Pyspark, Greykite, Pandas, Plotly, Hugging Face Transformers, PytimeTK

Language and IDE - Python (Jupyter Notebook)

3. Analysis Charts:

newplot (9) newplot (10) newplot (11) newplot (12) newplot (13) newplot (14) newplot (15) newplot (17)

4. Model Used and Evaluation Metrics:

  1. Link for PatchTsm Transformers Model (Hugging Face) - https://huggingface.co/docs/transformers/en/model_doc/patchtsmixer
  2. PatchTSM model explanation link - https://arxiv.org/pdf/2306.09364

Evaluation Metrics:

  • Epochs: 25
  • Train Loss (MSE): 0.1123
  • Train MAE: 0.0052
  • Validation Loss (MSE): 0.0061
  • Validation MAE: 0.0036

5. Links:

NASDAQ Analysis File link (Full view) - https://colab.research.google.com/drive/1-q0N1KlCwJrgYKR4TTnzrWDRsh6U2WuC (view in incognito or Microsoft Edge)

About

Nasdaq Time Series Forecasting using PatchTSM Transformers model (Hugging Face)

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published