Financial risk prediction using decision tree with Twitter data and compare prediction accuracy with random forest

To use tweets about finance to help decide on a financial choice. To forecast if financial tweets are positive or negative, we use a new Decision Tree that is better at getting the right results compared to Random Forest. The proposed way of doing things looks at 10 examples in each of two groups. W...

Full description

Saved in:
Bibliographic Details
Main Authors: Sudheer Kumar Reddy, Vajrala, Priyadarsini, P. S. Uma
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
Decision trees
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:To use tweets about finance to help decide on a financial choice. To forecast if financial tweets are positive or negative, we use a new Decision Tree that is better at getting the right results compared to Random Forest. The proposed way of doing things looks at 10 examples in each of two groups. We want to make sure we have enough power to detect any differences between the groups, so we aim for an 80% chance of finding a difference if one exists. We also set a significance level of 0. 05, which means we want to be fairly confident in our findings. The Novel Decision Tree is more accurate (96. 4%) than the Random Forest (92. 7%) in predicting the sentiments of financial tweets. The Novel Decision Tree Algorithm is much better than the Random Forest at guessing the feelings in financial tweets.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0197565