Classification of house buildings based on land size using the K-nearest neighbor algorithm

Abstract

Residential is housing and areas built in a balanced manner with a certain composition in single houses, simple houses, medium-sized houses, and luxurious houses or forms of flats. When building a large, small, or medium-sized residential house, you must pay attention to the size of the land building area and the building area of the house you want to design. If you ignore this, it will impact building the house, such as the plans that have been made are not suitable. The problem that is quite important from various problems exists in densely populated settlements where the land area is minimal and sequentially will impact the building area, where each area will not fit indoors. The standard for broad needs in building simple housing in Indonesia has been regulated in SNI 03-1733-2004, the minimum area for a simple house is 36m2 or 9m2 per person (Indonesian National Standard Agency 2004). This study uses the K-Nearest Neighbor Algorithm to classify house buildings based on land size. The results of classification testing using the K-Nearest Neighbor Algorithm on the WEKA application have the highest accuracy, precision, and recall values, namely testing on a dataset of 80% (66 data) from training data and 20% (16 data) from testing data with accuracy values. 100% precision, 1,000 precision and 1,000 recall.