MARCH 15, 2022 (NewsRx) — By a News Reporter – Staff News Editor at Daily Insurance News — Findings from risk management research are discussed in a new report. According to reports from Western College by NewsRx reporters, the research said, “Trading in stocks has tremendous importance not only as a profession, but also as a source of income for individuals.”
The editors got a citation from the research of Western College“Many investment account holders use their portfolio appreciation (in the form of a combination of stocks or indices) as income for their retirement years, betting primarily on low-risk stocks or indices/ low volatility. However, every stock-based investment portfolio has an inherent risk of losing money due to negative upside and crash. This study presents a new technique for predicting such negative events rare in financial time series (e.g. a decline in the S&P 500 by a certain percentage in a designated period of time. We use a time series of approximately seven years (2517 values) of stocks in the S&P 500 index with publicly available features: high, low and close price (HLC).We use a Siamese-like neural network for pattern recognition in images, followed by a distribution tion of image similarity primed to predict rare events related to financial market analysis. Extending to the literature on rare event classification and stochastic modeling in financial analysis, the proposed method uses a sliding window to store input characteristics as tabular data (HLC prices), creates an image of the time series window and then uses the feature vector of a pre-trained convolutional neural network (CNN) to mine pre-event images and predict rare events.
According to the news editors, the research concluded, “This research not only indicates that our proposed method is capable of distinguishing event images from non-event images, but more importantly, the method is effective even when only limited and severely unbalanced data are available.”
For more information on this research, see: A Novel Implementation of Siamese Type Neural Networks in Predicting Rare Fluctuations in Financial Time Series. Risks, 2022.10(39):39. (Risks – http://www.mdpi.com/journal/risques). The Risks editor is MDPI SA.
Our journalists inform that additional information can be obtained by contacting Treena Basu, Mathematics department, Western College, Los Angeles, CA 90041, United States. Other authors of this research include Olaf Menzer, Joshua Neighborhood, Indranil Sen Gupta.
(Our reports provide factual information on research and discoveries from around the world.)