Adapting speech models for stock price prediction

Frederic Voigt; Jose  Alcarez Calero; Keshav Dahal; Qi Wang; Kai Von Luck; Peer Stelldinger

doi:10.1109/ICCCMLA63077.2024.10871633

Adapting speech models for stock price prediction

Frederic Voigt, Jose Alcarez Calero, Keshav Dahal, Qi Wang, Kai Von Luck, Peer Stelldinger

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

5 Downloads (Pure)

Abstract

Large language models (LLMs) have demonstrated remarkable success in the field of natural language processing (NLP). Despite their origins in NLP, these algorithms possess the theoretical capability to process any data type represented in an NLP-like format. In this study, we use stock data to illustrate three methodologies for processing regression data with LLMs, employing tokenization and contextualized embeddings. By leveraging the well-known LLM algorithm Bidirectional Encoder Representations from Transformers (BERT) [1], we apply quantitative stock price prediction methodologies to predict stock prices and stock price movements, showcasing the versatility and potential of LLMs in financial data analysis.

Original language	English
Title of host publication	2024 IEEE 6th International Conference on Cybernetics, Cognition and Machine Learning Applications (ICCCMLA)
Publisher	IEEE
Number of pages	8
ISBN (Electronic)	9798331505790
ISBN (Print)	9798331505806
DOIs	https://doi.org/10.1109/ICCCMLA63077.2024.10871633
Publication status	Published - 11 Feb 2025

Keywords

finance
quantitative stock price prediction
natural language processing
stock movement prediction
fintech
machine learning
large language models

Access to Document

10.1109/ICCCMLA63077.2024.10871633

2024 08 05 Voigt et al Models accepted - for PureAccepted author manuscript, 1.16 MBLicence: CC BY 4.0

Cite this

@inproceedings{c9bb3afa5855419d8b21d50ef283e710,

title = "Adapting speech models for stock price prediction",

abstract = "Large language models (LLMs) have demonstrated remarkable success in the field of natural language processing (NLP). Despite their origins in NLP, these algorithms possess the theoretical capability to process any data type represented in an NLP-like format. In this study, we use stock data to illustrate three methodologies for processing regression data with LLMs, employing tokenization and contextualized embeddings. By leveraging the well-known LLM algorithm Bidirectional Encoder Representations from Transformers (BERT) [1], we apply quantitative stock price prediction methodologies to predict stock prices and stock price movements, showcasing the versatility and potential of LLMs in financial data analysis.",

keywords = "finance, quantitative stock price prediction, natural language processing, stock movement prediction, fintech, machine learning, large language models",

author = "Frederic Voigt and {Alcarez Calero}, Jose and Keshav Dahal and Qi Wang and {Von Luck}, Kai and Peer Stelldinger",

year = "2025",

month = feb,

day = "11",

doi = "10.1109/ICCCMLA63077.2024.10871633",

language = "English",

isbn = "9798331505806",

booktitle = "2024 IEEE 6th International Conference on Cybernetics, Cognition and Machine Learning Applications (ICCCMLA)",

publisher = "IEEE",

address = "United States",

}

TY - GEN

T1 - Adapting speech models for stock price prediction

AU - Voigt, Frederic

AU - Alcarez Calero, Jose

AU - Dahal, Keshav

AU - Wang, Qi

AU - Von Luck, Kai

AU - Stelldinger, Peer

PY - 2025/2/11

Y1 - 2025/2/11

N2 - Large language models (LLMs) have demonstrated remarkable success in the field of natural language processing (NLP). Despite their origins in NLP, these algorithms possess the theoretical capability to process any data type represented in an NLP-like format. In this study, we use stock data to illustrate three methodologies for processing regression data with LLMs, employing tokenization and contextualized embeddings. By leveraging the well-known LLM algorithm Bidirectional Encoder Representations from Transformers (BERT) [1], we apply quantitative stock price prediction methodologies to predict stock prices and stock price movements, showcasing the versatility and potential of LLMs in financial data analysis.

AB - Large language models (LLMs) have demonstrated remarkable success in the field of natural language processing (NLP). Despite their origins in NLP, these algorithms possess the theoretical capability to process any data type represented in an NLP-like format. In this study, we use stock data to illustrate three methodologies for processing regression data with LLMs, employing tokenization and contextualized embeddings. By leveraging the well-known LLM algorithm Bidirectional Encoder Representations from Transformers (BERT) [1], we apply quantitative stock price prediction methodologies to predict stock prices and stock price movements, showcasing the versatility and potential of LLMs in financial data analysis.

KW - finance

KW - quantitative stock price prediction

KW - natural language processing

KW - stock movement prediction

KW - fintech

KW - machine learning

KW - large language models

U2 - 10.1109/ICCCMLA63077.2024.10871633

DO - 10.1109/ICCCMLA63077.2024.10871633

M3 - Conference contribution

SN - 9798331505806

BT - 2024 IEEE 6th International Conference on Cybernetics, Cognition and Machine Learning Applications (ICCCMLA)

PB - IEEE

ER -

Adapting speech models for stock price prediction

Abstract

Keywords

Access to Document

Fingerprint

Cite this