A research on the comparison of classification algorithm in finance

Abstract

Nowadays, data is developing with technology. With developments in the Internet and technology, it makes a warehouse that changes the data more rapidly and in volume. This situation increases the importance of data mining every day. The aim of this study is to examine classification algorithms which are the sub-subject of data mining. In the analysis part of the study, it is aimed to compare the algorithms with the accuracy ratio by using the financial data. As a result of the study conducted on stock marketing, the classification of enterprises of decision tree models with artificial neural networks has also been found to be more successful than other methods.

FITTING Data Mining Settings for Ranking Seed Lots

“…Data mining is a way of finding unknown information in large datasets (Eti & İnel, 2019). With the development of computers and advances in algorithm innovation, as well as greater data availability, mining methods have undergone significant technological advances (Artrith et al, 2021).…”

“…In this sense, machine learning techniques and statistical analyses have been developed to analyze large data sets. Machine learning can be divided into two categories: supervised and unsupervised (Eti & İnel, 2019).…”

“…In large datasets, more exotic and deep algorithms, such as neural networks, should be adopted, while in smaller datasets, more classical techniques, such as linear regression and decision trees, are recommended, taking care to adapt each approach to data characteristics (Nichols et al, 2018). Decision tree algorithms create a sequence of rules to increase knowledge gain, enabling classifying data from a model composed of branches and leaves (Eti & İnel, 2019). For this technique, there are some models, such as ID3, C4.5, and J48, which is an evolution of the latter (Joshuva et al, 2020).…”