The paper explores the intraday price movement of stocks as an image classification problem using a CNN-based model. The study demonstrates a high-level relationship between the first hour of trading and the close, with the algorithm adequately separating between opposing classes. Investing according to the algorithm's predictions outperformed all alternative constructs except for the theoretical maximum. To support their thesis, several additional tests were conducted, highlighting the suitability of computer vision techniques for studying financial markets and predicting stock price movements. The study also addresses overfitting by adding a dropout layer that discards 20% of random units prior to classification. The paper discusses various financial and economic variables proposed as potential predictors, including valuation ratios, interest rates and spreads, macroeconomic variables such as inflation and industrial production, and psychology in asset pricing. While these papers contributed to fundamental understanding, they had limited use for predicting individual stock price movements. Recent advancements in artificial intelligence have entered the field of financial markets, with some examples combining pipelines of 1-dimensional CNN models with bi-directional LSTM units or using CNN models on limit order books. The paper concludes that asset-trading can be considered an image classification problem done on an in-sample dataset. The author suggests that technological innovation may diminish the size of investment effects while emphasizing that a smaller portfolio is more manageable. Matej Steinbacher graduated in economics from the University of Maribor and continued post-graduate studies at the Faculty of Economics at the University of Ljubljana focused on network models in economics and finance.
- - The paper uses a CNN-based model to explore intraday price movement of stocks as an image classification problem.
- - The study shows a high-level relationship between the first hour of trading and the close, with the algorithm separating opposing classes effectively.
- - Investing according to the algorithm's predictions outperformed all alternative constructs except for the theoretical maximum.
- - Several additional tests were conducted to support their thesis, highlighting the suitability of computer vision techniques for studying financial markets and predicting stock price movements.
- - The study addresses overfitting by adding a dropout layer that discards 20% of random units prior to classification.
- - Various financial and economic variables proposed as potential predictors are discussed, including valuation ratios, interest rates and spreads, macroeconomic variables such as inflation and industrial production, and psychology in asset pricing.
- - Recent advancements in artificial intelligence have entered the field of financial markets, with some examples combining pipelines of 1-dimensional CNN models with bi-directional LSTM units or using CNN models on limit order books.
- - Asset-trading can be considered an image classification problem done on an in-sample dataset.
- - Technological innovation may diminish the size of investment effects while emphasizing that a smaller portfolio is more manageable.
This paper talks about using a computer program to predict how the stock market will change during the day. The program is really good at telling if the stock prices will go up or down. People who used this program to invest their money made more money than people who didn't use it. The scientists did some extra tests to make sure their program was working well and talked about other things that might help predict stock prices. They also added something to make sure their program wasn't just memorizing things instead of really understanding them. Some people are using new technology like this to help them invest in stocks, but it's important to remember that having a smaller amount of investments can be easier to manage.
Definitions- CNN-based model: A type of computer program that can look at pictures and understand what they show.
- Intraday price movement: How much the price of something changes during one day.
- Algorithm: A set of instructions for a computer program to follow.
- Overfitting: When a computer program is too good at remembering specific examples instead of understanding the bigger picture.
- Dropout layer: A part of a computer program that randomly ignores some information so the program doesn't get too good at memorizing things.
- Valuation ratios: Numbers that show how expensive or cheap something is compared to how much money it makes.
- Interest rates and spreads: How much money you have to pay back when you borrow money from someone else.
- Macroeconomic variables: Big-picture things like how many
Exploring Intraday Price Movement of Stocks as an Image Classification Problem
In a recent research paper, Matej Steinbacher explored the intraday price movement of stocks as an image classification problem using a CNN-based model. The study demonstrated a high-level relationship between the first hour of trading and the close, with the algorithm adequately separating between opposing classes. Investing according to the algorithm's predictions outperformed all alternative constructs except for the theoretical maximum.
The Model
To support their thesis, several additional tests were conducted, highlighting the suitability of computer vision techniques for studying financial markets and predicting stock price movements. The model used in this study was based on convolutional neural networks (CNNs). A dropout layer was added to address overfitting by discarding 20% of random units prior to classification.
Predictors
Various financial and economic variables were proposed as potential predictors including valuation ratios, interest rates and spreads, macroeconomic variables such as inflation and industrial production, and psychology in asset pricing. While these papers contributed to fundamental understanding, they had limited use for predicting individual stock price movements.
Recent Advancements in AI
Recent advancements in artificial intelligence have entered the field of financial markets with some examples combining pipelines of 1-dimensional CNN models with bi-directional LSTM units or using CNN models on limit order books. The paper concluded that asset-trading can be considered an image classification problem done on an in-sample dataset.
Conclusion & Suggestions
The author suggested that technological innovation may diminish the size of investment effects while emphasizing that a smaller portfolio is more manageable. This research paper provides insight into how computer vision techniques can be used to accurately predict stock prices movements which could potentially lead to improved returns for investors who are able to capitalize on this new technology quickly enough before it becomes widely adopted by other traders/investors in the market place.