Claim Missing Document
Check
Articles

Found 4 Documents
Search
Journal : Jurnal Mandiri IT

Comparison of naïve bayes and KNN for herbal leaf classification Nugroho, Bangkit Indarmawan; Khusni, Muhammad Wazid; Ananda, Pingky Septiana; Gunawan, Gunawan
Jurnal Mandiri IT Vol. 13 No. 1 (2024): July: Computer Science and Field
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mandiri.v13i1.297

Abstract

This study aims to compare the effectiveness of two classification algorithms, namely Naïve Bayes Classifier and K-Nearest Neighbor (KNN), in classifying herbal leaves. This research design uses a quantitative approach with experimental analysis and model validation. The dataset consisted of images of papaya leaves, pandanus, cat's whiskers, and betel nut taken in different lighting conditions. The methodology includes pre-processing of data by converting images into grayscale, feature extraction using Gray Level Co-occurrence Matrix (GLCM), and application of Naïve Bayes and KNN algorithms. The main results showed that KNN achieved 90.00% accuracy with precision, recall, and F1-score of 88.33% respectively, higher than Naïve Bayes which had 82.50% accuracy, 81.46% precision, 85.83% recall, and 82.27% F1-score. In conclusion, KNN is superior in the classification of herbal leaves to Naïve Bayes, although it requires a longer computational time. Further research is recommended to optimize algorithm parameters and explore the integration of deep learning techniques to improve classification accuracy and efficiency.
Application of centroid and geometric mean methods for face recognition Nugroho, Bangkit Indarmawan; Khasanah, Apriliani Maulidya; Arif, Zaenul; Gunawan, Gunawan
Jurnal Mandiri IT Vol. 13 No. 1 (2024): July: Computer Science and Field
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mandiri.v13i1.300

Abstract

Face recognition is one of the most important areas in artificial intelligence and image processing, with wide applications from attendance system security to human-computer interaction. This study aims to overcome the difficulties in classifying student faces in an academic environment by applying and comparing centroid and geometric mean methods. Student face data was collected and processed through conversion to grayscale, pixel intensity normalization, and statistical analysis using both methods. The results showed that both methods had the same performance with 70% accuracy, 75% precision, 60% recall, and 66.67% F1-score. The application of this method can improve the efficiency and accuracy of attendance management and security in the campus environment, especially for institutions with limited resources.
Application of WASPAS method in determining the best flour for nastar making Nugroho, Bangkit Indarmawan; Dewi, Errika Mutiara; Kurniawan, Rifki Dwi; Gunawan, Gunawan
Jurnal Mandiri IT Vol. 13 No. 1 (2024): July: Computer Science and Field
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mandiri.v13i1.303

Abstract

This study explores the use of the Weighted Aggregated Sum Product Assessment (WASPAS) Method in selecting the best wheat flour for pineapple cake production. The aim of this study is to develop a more systematic and quantitative approach in assessing flour quality, provide useful guidance for pineapple cake producers and enrich the academic literature in the field of food science and food technology. This study used quantitative methodology data analysis and model validation with WASPAS, aimed at overcoming the challenge of selecting the best wheat flour for pineapple cake making. Results showed that the WASPAS method was effective in identifying the best flour, with Bungasari Hana Emas flour obtaining the highest WASPAS score of 0.952863, followed by the Falcon Hijau with a score of 0.931373. This score indicates the optimal balance between cost and quality. The study emphasizes the importance of objective decision-making tools in the food industry, suggesting that such an approach can significantly improve product quality and production efficiency.
Applying certainty factor method to identify diseases in rice plants Nugroho, Bangkit Indarmawan; Miftakhuddin, Ahmad; Syefudin, Syefudin; Gunawan, Gunawan
Jurnal Mandiri IT Vol. 13 No. 1 (2024): July: Computer Science and Field
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mandiri.v13i1.310

Abstract

Rice (Oryza Sativa L) is the most important food crop in the world after wheat and corn, as well as the main source of protein for most of the world's population, especially in Asia. The Save Swamps for Prosperous Farmers (Serasi) program in Central Java Territory cannot run well considering the tall capacity of existing rice agriculturists to bargain with bugs and maladies of the rice they plant, so it is essential to make a device within the frame of an master framework for diagnosing rice plant infections.  For this reason, it is very important to be aware of the factors that influence production levels. Disease is one of the most detrimental factors in rice production, where many losses are caused by disease. Each of these diseases generally shows symptoms of the disease suffered before it reaches a more severe and widespread stage, these symptoms can be recognized by carrying out a diagnosis first. This can be done using an expert system. In this research, an expert system was utilized which was made utilizing the certainty figure strategy, with a test of 25 ranchers within the West Tegal Area, Tegal City. From the comes about of the inquire about carried out, it was concluded that with this framework the level of exactness obtained using the posttest contains a esteem of 100%, in other words the framework encompasses a decently tall level of accuracy.