Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Automatic market analysis method based on multi-mode learning

A market analysis and multi-modal technology, applied in marketing and other directions, can solve problems such as risk increase, market change grasp, market change error prediction, etc., to improve performance and accuracy, reduce risk, and have strong applicability Effect

Active Publication Date: 2015-01-28
SUZHOU CHENCHUAN COMM TECH
View PDF2 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most of the current algorithmic trading methods tend to focus on one side of the market, such as price or related news reports, and only consider a single factor and take it into consideration as a decisive factor. This method leads to the inability of automatic trading algorithms to obtain comprehensive market information. , so that it is difficult to grasp the changes in the market, which will lead to inaccurate forecasts of market changes, and even make wrong forecasts of market changes, which in turn will increase the risk

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0022] Embodiment 1, a kind of automatic market analysis method based on multimodal learning, first trains the learning device, and then utilizes the trained learning device to predict the market in actual use; the method for the training of the learning device is: first Collect information of different modalities such as market data or market description text information and mark the information; then use the multi-instance generation method of market data features and text features to convert the underlying features into the form of multi-instance packages; finally adopt the Using a variety of multi-instance multi-label learning methods of different modalities to fuse data and perform multi-label learning.

Embodiment 2

[0023] Embodiment 2, in the automatic market analysis method based on multimodal learning described in Embodiment 1: the specific steps of the training of the learning device are as follows:

[0024] Step 100, collect various modal information of the market, and manually mark the collected objects;

[0025] Step 101, convert the collected underlying features of market information into a multi-instance package representation through the multi-instance generation method: {(x, t), y}, where the media object is marked as x, and the corresponding other modal information is recorded as t , manually marked as y;

[0026] Step 102, use the training model M to train the collected data to obtain relevant model parameters: marker generation sub-model parameters α, β y ;Market direct market eigenmode information generation sub-model parameter β c ;Other modal information generation sub-model parameters β t And the multimodal input latent variable controls the model parameter η.

[002...

Embodiment 3

[0032] Embodiment 3, in the automatic market analysis method based on multimodal learning described in embodiment 1 and embodiment 2: the specific steps of the training of the learning device are as follows: the steps of using the learning device are as follows:

[0033] Step 200, collect test market data characteristics (if there are other modal data, also collect);

[0034] Step 201, transform the underlying characteristics of the market conditions into the representation form {(x)} or {(x, t)} of the multi-instance package through the multi-instance generation method;

[0035] Step 202, use the training model M to process the new market feature I, and output the prediction mark y.

[0036] The generative probability model modeling method of the training model M is:

[0037] (1) (marker-theme submodel part) Let the market quotations be generated by the theme model, where the mark y is determined by the parameter α through the Latent Dirichlet Allocation (LDA) submodel and t...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an automatic market analysis method based on multi-mode learning. The automatic market analysis method comprises the steps that firstly, a learning device is trained, and then in practical use, the trained learning device is used for predicting a market. A method for training the learning device comprises the steps that firstly, different-mode information of the market is collected and labeled; then, a multi-example generating method of market data features and character features is used for converting low-level features into a multi-example-packet format; finally, a multi-example multi-label learning method capable of utilizing various modes is adopted to conduct fusion processing on the data, and multi-label learning is conducted. According to the automatic market analysis method based on multi-mode learning, by acquiring multiple pieces of side information of the market, market changes can be depicted more comprehensively and market changes can be predicted more accurately. According to the method for conducting market prediction by using the multi-mode information, the different-mode data information can be used in the implementation process, adaptability is high and effects are good.

Description

technical field [0001] The present invention belongs to the field of automated market analysis methods, in particular to an automated market analysis method based on multimodal learning. Background technique [0002] Algorithmic trading, which uses a computer platform to input trading orders by executing a pre-set trading strategy, has become one of the mainstream trading methods at present. Considering that the composition of the market microstructure is restricted by various constraints, that is, the micro market conditions are related to multiple different sources of information. However, most of the current algorithmic trading methods tend to focus on one side of the market, such as price or related news reports, and only consider a single factor and take it into consideration as a decisive factor. This method leads to the inability of automatic trading algorithms to obtain comprehensive market information. , so that it is difficult to grasp the changes in the market, w...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06Q30/02
Inventor 詹德川周尚晨
Owner SUZHOU CHENCHUAN COMM TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com