Comparative performance analysis of random forest and logistic regression algorithms Rastgele orman ve lojistik regresyon algoritmalanmn karşllaştirmali performans analizi
5th International Conference on Computer Science and Engineering, UBMK 2020, Diyarbakır, Türkiye, 9 - 10 Eylül 2020, ss.25-30, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/ubmk50275.2020.9219478
- Basıldığı Şehir: Diyarbakır
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.25-30
- Anahtar Kelimeler: Bank Marketing Data, Logistic Regression, MATLAB, Random Forest, WEKA
- Maltepe Üniversitesi Adresli: Evet
Özet
Today, banks are trying to meet the needs of their existing customers with the marketing activities they do in digital media. It is known to produce statistical results in order to be able to predict the behavior of customers in artificial intelligence applications by storing large-scale data obtained through marketing studies. In this study, performance comparison between random forest and logistic regression algorithms was made by using real banking marketing data that includes the characteristics of customers. In addition, these algorithms were run on WEKA, Google Colab and MATLAB platforms to compare performance on different platforms. At the end of the study, the most successful result obtained with 94.8% accuracy, 93.9% sensitivity, 94.8% recall, 94.4% fl-score and 98.7% AUC value was achieved by random forest algorithm on WEKA platform. In addition, it has been shown that the obtained performance values produce better results compared to similar studies.