Finassociations: Investment Recommendation Application with Association Rule Mining
Authors: Elif KARTAL, M. Erdal BALABAN, and Zeki OZEN
Abstract: Finassociations is a flexible and interactive investment recommendation application that enables app users to discover financial relations between currencies, cryptocurrencies, and stocks through Association Rule Mining (ARM). ARM is a technique generally used for market basket analysis; also, it has a variety of applications in other areas since it facilitates the discovery of associations in transactions. The app allows users to select investment instruments, retrieve historical data from Yahoo Finance, and analyze bilateral correlations. In the Association Rule Mining section, users can set parameters like rate of change, confidence, and support to generate and visualize association rules, with results available for export in various formats.
Full Description: Finassociations is a flexible and interactive investment recommendation application that enables app users to discover financial relations between currencies, cryptocurrencies, and stocks through Association Rule Mining (ARM). ARM is a technique generally used for market basket analysis; also, it has a variety of applications in other areas since it facilitates the discovery of associations in transactions. Finassociations Application has two main parts: Data Selection and Association Rule Mining.
In the Data Selection section, the user is expected to choose among the investment instruments is interested in with a date range. This section automatically retrieves data for the selected investment instruments from Yahoo Finance. This data can be displayed in tabular form. Bilateral correlations between investment instruments are presented numerically and visually. In addition, with the help of a line graph, the change in the selected investment instrument can be displayed. Clicking on the “Next” button activates the Association Rule Mining Section tab. In this section, the user can select ARM-related parameters such as the rate of change (%), the minimum confidence, the minimum support, and the minimum rule length. The “Get Rules” button returns the analysis results. The rate of change, the rate of change status, and association rules will be shown in tables. The user can save or print the analysis results in different file formats such as CSV, Excel, or PDF. The rules are also presented to the user visually. Finally, the top ten most frequently recurring items in the given date range are shown in a bar chart. There is also a Help Section that introduces the application, some parameters associated with the ARM and gives information on how to read the rules obtained. Furthermore, users can request more financial instruments (currency, cryptocurrency, or stock market data) to analyze if needed.
Finassociations has been developed as part of a scientific study entitled 'Investment analytics using association rule mining (Finassociations)' by Kartal, E., Balaban, M. E., and Ă–zen, Z., and has been accepted for publication in the International Journal of Computational Economics and Econometrics Special Issue on 'Economic Analysis and the Current Real-World Situation: Exploring New Trends in Applied Economics,' in honor of Prof. George Agiomirgianakis. Since the app has been developed within the scope of scientific research, the information obtained from Finassociations should be used for informational and research purposes only. It is not intended to provide investment or other professional financial advice.
Shiny app: https://erdalbalaban.shinyapps.io/finassociations/
Repo: GitHub - elifkrtl/finassociations: Finassociations: Investment Recommendation Application with Association Rule Mining
Thumbnail:
Full image:
