Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : INFOKUM

Clustering Potato Seeds Using DBSCAN Algorithm in Optimizing Sales N P Dharshinni; Andre Putra Persada Tarigan
INFOKUM Vol. 10 No. 1 (2021): Desember, Data Mining, Image Processing, and artificial intelligence
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (817.013 KB)

Abstract

Potato seed shop CV. ASAP sells various types of potato seeds which are quite trusted by potato farmers. The payment methods used so far are cash and credit payments, but the shop owner never has a balanced stock so that customer demand is always not in accordance with the stock of seeds, resulting in the shop is at a loss. Based on the problems experienced by potato farmers, it is necessary to group potato seed sales data using the DBSCAN clustering algorithm. The DBSCAN algorithm groups sales data into several cluster groups based on epsilon values and minimal points, besides that this algorithm can detect noise in data grouping so that the grouping results obtained are better than other algorithms. The purpose of this study was to apply the DBSCAN algorithm to potato seedling sales data to obtain results in knowing the types of potato seeds that were most purchased by customers from potato seed sales with cash payment methods and credit payments. The results of the application of the DBSCAN algorithm found that the most purchased by customers using the cash payment method were the type of spread seeds, class D potato seeds, class A potato seeds, and class C potato seeds with the test value of epsilon 1,2 and Minimum points 1 while the sale of seeds Potatoes, the most purchased by a customer using credit payment method are the type of potato produced by class C and potato spread seeds with the test value of Epsilon 2 and Minimum points 1.
IMPROVEMENT OF DIGITAL IMAGE USING A COMBINATION OF ALPHA TRIMMED MEAN FILTER AND ARITHMETIC MEAN FILTER Insidini Fawwaz; N P Dharshinni; Irfan Hindrawan
INFOKUM Vol. 10 No. 02 (2022): Juni, Data Mining, Image Processing, and artificial intelligence
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (944.539 KB)

Abstract

The development of technology at this time causes the provision of information to increase through social media. Many social media users convey information by including digital images. Digital images are very important in conveying the accuracy of information. However, digital images often experience various disturbances, such as decreased pixel quality, less sharpness, blurring, and the appearance of noise in the image. Noise contained in the image causes a decrease in image quality. Image degradation can be caused by uneven light intensity and can also be caused by dirt adhering to the camera lens. There are various types of noise found in digital images, including Salt And Pepper Noise, Speckle Noise, and Rayleigh Noise. There are many filtering methods that can improve digital images from noise interference. Some of them are the Mean Filter method, Geometric Mean Filter, Harmonic Mean Filter, Arithmetic Mean Filter, Median Filter, Midpoint filter, Alpha Trimmed Mean Filter and so on. Based on the research conducted, the combination of the Alpha Trimmed Mean Filter and Arithmetic Mean Filter methods can reduce Salt and Pepper noise, Speckle noise and Rayleigh noise better than the Alpha Trimmed Mean Filter and Arithmetic Mean Filter methods based on the MSE, RMSE and PSNR parameters.