Forecasting educated unemployed people in Indonesia using the Bootstrap Technique

Forecasting is an essential analytical tool used to make future predictions based on preliminary data. However, the use of small sample sizes during analysis provides inaccurate results, known as asymptotic forecasting. Therefore, this study aims to analyze the unemployment rate of educated people in Indonesia using the bias-corrected forecasting bootstrap technique. Data were collected from a total of 30 time series of educated unemployed from 2015 to 2019 using the bias-corrected boot-strap technique and determined using the interval prediction method. The bootstrap replication used is at intervals of 100, 250, 500, 1000, and 10000. The results obtained using the R program showed that the boot-strap technique provides consistent forecasting results, better accuracy, and unbiased estimation. Moreover, the results also show that for the next 10 periods, the number of educated unemployed people in Indonesia is projected to decline. The bootstrap coefficient also tends to decrease with an increase in the number of replications, at an average of 0.958. The interval prediction is also known to be smooth, along with a large number of bootstrap replications.