Comparing Random Forest Algorithm and Support Vector Machine for Predicting the Level of Satisfaction with Flights

SOESANTIO, WAHONO (2022) Comparing Random Forest Algorithm and Support Vector Machine for Predicting the Level of Satisfaction with Flights. Other thesis, Universitas Katholik Soegijapranata Semarang.

[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-COVER_a.pdf

Download (569kB) | Preview
[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-BAB I_a.pdf

Download (299kB) | Preview
[img] Text
18.K1.0045-WAHONO SOESANTIO-BAB II_a.pdf
Restricted to Registered users only

Download (244kB)
[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-BAB III_a.pdf

Download (357kB) | Preview
[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-BAB IV_a.pdf

Download (625kB) | Preview
[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-BAB V_a.pdf

Download (501kB) | Preview
[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-BAB VI_a.pdf

Download (230kB) | Preview
[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-DAPUS_a.pdf

Download (260kB) | Preview
[img]
Preview
Text
18.K1.0045-WAHONO SOESANTIO-LAMP_a.pdf

Download (286kB) | Preview

Abstract

This study aims to find out which method is better and more effective so that it can be seen the difference between the two algorithms in predicting the level of passenger satisfaction with public services on the plane which will be displayed using a graph. For the completion process, what I did was try each algorithm, the first thing I did was to try the random forest algorithm then when I tried it, I kept looking for the TP, TN, FP, FN values to generate prediction scores, accuracy, recall, f-1 scores. then compared with the support vector machine algorithm with more or less the same steps by looking for a true positive, true negative, false negative, false positive then looking for predictive value, accuracy, recall, the f-1 score then displayed on a graph to see which one is better. Both algorithms using a cross-validation split for split data which aims to divide the data into training and testing data and make it easier to calculate accuracy. For the use of split data used with a comparison of 70% training data and 30% testing data from the whole data. The final result can be seen from the graph of each algorithm that has shown the prediction value, recall, precision, f-1 score and from there it can be concluded which method is better and the differences from each existing algorithm. To get the accuracy results obtained from the determination of the confusion matrix in the form of TP, TN, FP, and FN.

Item Type: Thesis (Other)
Subjects: 000 Computer Science, Information and General Works
Divisions: Faculty of Computer Science > Department of Informatics Engineering
Depositing User: mr AM. Pudja Adjie Sudoso
Date Deposited: 26 Oct 2022 09:28
Last Modified: 26 Oct 2022 09:28
URI: http://repository.unika.ac.id/id/eprint/30029

Actions (login required)

View Item View Item