Online ISSN: 2515-8260

Keywords : Support vector machines


A Brief Study on the application of SVM Algorithm for Asset Price Prediction and Portfolio Optimization with respect to Risk and Return

Sabarinathan .; A. Muhammad Raheel Basha; J. Dinesh; U. Thilak; R. Muruganandham; S. Vanitha

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 2, Pages 4991-4997

Portfolio Optimization is to evolve models to compute an optimal proportion of capital for investing with respects to the assets in the portfolio. Portfolio optimization covers a wide range of financial assets, such as stocks, funds, bonds, commodities, currencies and loans, whereas similar concepts and ideas are also applicable to non-financial portfolios. Asset price prediction is an important challenge in portfolio optimization. This project utilizes Support Vector Machines, a Machine learning algorithm for asset price prediction. SVM is very accurate and gives better results compared to other techniques. This project is mainly concentrated on predicting asset price followed by portfolio optimization considering the risk and return associated with each and every asset using R programming.

Text Mining Based on Tax Comments as Big Data Analysis Using XGBOOST and Feature Selection

RAVI KUMAR B.CHAWAN, KORIVI VAMSHEE KRISHNA, SIRIKONDA VAMSHI KRISHNA

European Journal of Molecular & Clinical Medicine, 2017, Volume 4, Issue 1, Pages 150-157

With the quick improvement of the Internet, enormous information has been applied in a lot of use.
Be that as it may, there are regularly excess or unessential highlights in high dimensional information, so
include determination is especially significant. By building subsets with new highlights and utilizing AI
calculations including Xgboost and so on. To acquire early notice data with high dependability and constant by
applying large information hypothesis, systems, models and techniques just as AI strategies are the unavoidable
patterns later on. this examination proposed the fast choice of highlights by utilizing XGboost model in dispersed
circumstances can improve the Model preparing proficiency under conveyed condition.
GBTs model dependent on the inclination streamlining choice tree was superior to the next two models as far as
precision and continuous execution, which meets the necessities under the large information foundation. It runs
on a solitary machine, just as the conveyed preparing structures Apache Hadoop, Apache Spark.
We can utilize inclination plummet for our slope boosting model. On account of a relapse tree, leaf hubs produce
a normal inclination among tests with comparative highlights. Highlight determination is a basic advance in
information preprocessing and significant research content in information mining and AI assignments, for
example, order.