Determination of elderly card recipients in west Jakarta by modeling C4.5 algorithm using Python

Provision of various social protection programs, social assistance, social rehabilitation and family-based social services for the elderly, the program can create smart living indicators so that residents have a quality life. The Jakarta Elderly Card (KLJ) is a program to fulfill the basic needs of elderly citizens from the DKI Jakarta Provincial Government. Problems in the field are related to the desire of the elderly to get KLJ. Meanwhile, many residents still do not know the scheme or rules for collecting data on KLJ recipients. Currently, KLJ recipients are still using the 2015 Integrated Database (BDT), one way to avoid disbursing KLJ that is not on target, namely through data upgrades. Applying the classification of the data mining method, namely the C4.5 algorithm using 4 parameters, namely marital status, latest education, gender and age. These parameters are then converted using the One-Hot-Encode method by utilizing the dummys function so that new parameters are formed with the contents of the record in the form of binary values. The resulting pattern in determining the data for Jakarta Elderly Card recipients is 10 rules. The elderly with the status of KLJ (Receive) were 84 people, and those with the status of KLJ (Not Receive) were 115 people. Evaluation of classification results in determining Jakarta Elderly Card recipients using sample data with the help of python tools resulted in an accuracy rate of 88.5% on training data, and 91.6% on testing data.