Modeling of Money Supply Using LASSO Regression and Simulated Annealing

Trisha Magdalena Adelheid Januaviani , Sukono , Eman Lesmana , Kalfin

The money supply has several factors that affect the bill to non-residents, obligations to non-residents, net bill to the central government, bill to other financial institutions, bill to local governments, bill to companies not State- Owned Enterprises and bill to the private sector. The bills must be correlated, which means that they have multicollinearity. The way to overcome the multicollinearity problem in money supply is the Least Absolute Shrinkage and Selection Operator (LASSO) regression method. LASSO is a shrinking method which shrinks the β parameter to zero or becomes zero. The LASSO model is compared to the simulated annealing algorithm model by looking at the lowest Cp Mallow's value.

Volume 12 | Issue 6

Pages: 840-848

DOI: 10.5373/JARDCS/V12I6/S20201100