Method for forecasting stock prices using hybrid machine learning model that combines deep learning and classical statistical techniques for financial markets by investors to make informed decisions, involves predicting stock price for future time points with error metrics for accuracy assessment
	MEENU; RAUTELA G; GUPTA P; SINGH R; GUPTA R
	2024-10-19
专利权人	MEENU (MEEN-Individual) ; RAUTELA G (RAUT-Individual) ; GUPTA P (GUPT-Individual) ; SINGH R (SING-Individual) ; GUPTA R (GUPT-Individual)
申请日期	2024-10-19
专利号	IN202411079546-A
成果简介	NOVELTY - The method involves inputting several historical stock price data attributes including open, high, low, close and volume data points and relevant macroeconomic indicators. Several models are trained. The data is processed through a deep learning module. The module comprises a neural network model for capturing nonlinear dependencies. The data is processed through a classical statistical module by a computing device. The module comprises an Autoregressive Integrated Moving Average (ARIMA) model for identifying linear trends. The outputs from both modules are combined to generate a forecast of stock prices. The stock price for future time points is predicted with error metrics for accuracy assessment. The accuracy and reliability of stock price predictions are optimized. The deep learning module utilizes recurrent neural network (RNN) architecture to analyze time-series data. The classical statistical module comprises seasonal adjustments to accommodate periodic trends in stock prices. USE - Method for forecasting stock prices using hybrid machine learning model that combines deep learning and classical statistical techniques for financial markets by investors, analysts, and portfolio managers to make informed decisions. ADVANTAGE - The method improves forecasting accuracy by using the predictive capabilities of deep learning alongside the robustness of statistical techniques. The method discloses the hybrid model which demonstrated a remarkable accuracy rate of 100%, affirming the effectiveness of the integrative approach in improving forecasting precision and provides a hybrid machine learning framework that leverages the strengths of both deep learning and classical statistical techniques for stock price forecasting. The effectiveness of the hybrid approach is evaluated using accuracy metrics such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R-squared (R^2) values, alongside a comparative analysis against traditional forecasting methods to validate its effectiveness.
IPC 分类号	G06N-003/02 ; G06N-003/08 ; G06Q-040/04
国家	印度
专业领域	信息技术
语种	英语
成果类型	专利
文献类型	科技成果
条目标识符	http://119.78.100.226:8889/handle/3KE4DYBR/14839
专题	中国科学院新疆生态与地理研究所
作者单位	1.MEENU (MEEN-Individual) 2.RAUTELA G (RAUT-Individual) 3.GUPTA P (GUPT-Individual) 4.SINGH R (SING-Individual) 5.GUPTA R (GUPT-Individual)
推荐引用方式 GB/T 7714	MEENU,RAUTELA G,GUPTA P,et al. Method for forecasting stock prices using hybrid machine learning model that combines deep learning and classical statistical techniques for financial markets by investors to make informed decisions, involves predicting stock price for future time points with error metrics for accuracy assessment. IN202411079546-A[P]. 2024.

条目包含的文件
条目无相关文件。

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。