Enhancement of random forest by utilizing modified whale optimization algorithm / Ali Qjiram G. Pirzada, Renz Michael M. Leandicho. 6
By: Ali Qjiram G. Pirzada, Renz Michael M. Leandicho. 4 0 16 [, ] | [, ] |
Contributor(s): 5 6 [] |
Language: Unknown language code Summary language: Unknown language code Original language: Unknown language code Series: ; June 2023.46Edition: Description: 28 cm. 46 ppContent type: text Media type: unmediated Carrier type: volumeISBN: ISSN: 2Other title: 6 []Uniform titles: | | Subject(s): -- 2 -- 0 -- -- | -- 2 -- 0 -- 6 -- | 2 0 -- | -- -- 20 -- | | -- -- -- -- 20 -- | -- -- -- 20 -- --Genre/Form: -- 2 -- Additional physical formats: DDC classification: | LOC classification: | | 2Other classification:| Item type | Current location | Home library | Collection | Call number | Status | Date due | Barcode | Item holds |
|---|---|---|---|---|---|---|---|---|
| Book | PLM | PLM Filipiniana Section | Filipiniana-Thesis | QA76.9.a 43 P57 2023 (Browse shelf) | Available | FT7747 |
Undergraduate Thesis: (Bachelor of Science in Computer Science) Pamantasan ng Lungsod ng Maynila, 2023. 56
5
ABSTRACT: Machine learning (ML) have various applications, including the ability of software to predict and analyze results more correctly without explicit instructions, identify the best ways to automate tasks, enhance processes, and many other things. The Random Forest (RF) model has been proven to perform well and has applications in many different sectors, but current research suggests that there is still room for improvement. It is the most well-known and often used machine learning technique. There is still room for development with the RF model. In this paper, the researchers provided an optimization algorithm (WOA) to enhance and improve the accuracy of the Random Forest Algorithm on a UNSW-NB15 Intrusion detection dataset. It achieved an accuracy of 97.14% with the hybrid algorithm compared to the traditional algorithm of 94.79%. Furthermore, the recall scores for the proposed algorithm and traditional RF were 95.80% and 92.26% respectively, while the precision for MWOA-RF and traditional RF were equal at 1.000. It indicates that the suggested method performed better at correctly identifying positive cases and had a lower rate of false negatives recognized. Lastly, The F1-Score given by the MWOA-RF is 0.9785 compared to the F1-Score of the traditional RF, which is 0.9597, which signifies that the proposed MWOA-RF performs better for classification and is the better model for the two since its value is closer to 1. The results imply that MWOA-RF is faily more stable and with its properly tuned hyperparameters, is more suitable for carrying out classification tasks with huge datasets.
5
There are no comments for this item.