Estimating the unemployment rate using least square and conjugate gradient methods

  • Authors

    • Nur Syarafina Mohamed
    • Mustafa Mamat
    • Mohd Rivaie
    • Nur Hamizah Abdul Ghani
    • Norhaslinda Zull
    • Syazni Syoid
    2018-04-06
    https://doi.org/10.14419/ijet.v7i2.15.11360
  • Conjugate gradient method, Least square method, Unconstrained optimization, Workforce.

  • Abstract

    Unemployment rate is one of the major issues among Malaysian citizens. The unemployment rate indicates the percentage of the total workforce who are actively seeking employment and currently unemployed. In this paper, a data of unemployment rate of a state in Malaysia from year 2000 until 2015 is collected. The statistics data is extracted by Labour Force Survey Malaysia (LFSM) which was conducted monthly by using household approach targeted to working ages between 15 to 64 years old. An estimation data for year 2016 can be forecasted by using discrete least square method of numerical analysis and conjugate gradient method in unconstrained optimization. These methods have been chosen based on its simplicity and accuracy. The calculations are based on linear and quadratic models for each the method together with their errors. Results showed that the conjugate gradient method is comparable with the least square method.

     

  • References

    1. [1] Moyi AU, Leong WJ & Saidu I (2014), On the application of three-term conjugate gradient method in regression analysis. International Journal of Computer Applications 102, 1–4.

      [2] Chapra SC & Canale RP (2006), Numerical methods for engineers, McGraw-Hill.

      [3] Armstrong JS & Fildes R (1995), Correspondence on the selection of error measures for comparisons among forecasting methods. Journal of Forecasting 14, 67–71.

      [4] Makridakis S, Anderson S, Carbone R, Hibson M, Lewandoski R, Newton J, Parzen E & Wikler R (1984), The forecasting accuracy of major times series methods, John Wiley and Sons.

      [5] Narula S & Wellington J (1977), An algorithm for the minimum sum of weighted absolute errors regression. Communication in Statistics 6, 341–352.

      [6] Khoshgoftaar TM, Bhattacharyya BB & Richardson GD (1992), Predicting software errors during development using nonlinear regression models: A comparative study. IEEE Transactions on Reliability 41, 390–395.

      [7] Burden RL & Faires JD (2011), Numerical analysis: Interpolation and polynomial approximation, Cengage Learning.

      [8] Yuan G & Wei Z (2009), New line search methods for unconstrained optimization. Journal of the Korean Statistical Society 38, 29–39.

      [9] Mohamed NS, Mamat M, Susilawati F & Rivaie M (2016), A new coefficient of conjugate gradient method for nonlinear unconstrained optimization. Jurnal Teknologi 78, 131–136.

      [10] Mohamed NS, Mamat M & Rivaie M (2016) Solving a large scale nonlinear unconstrained optimization with exact line search direction by using new coefficient of conjugate gradient methods. AIP Conference Proceedings 1787, 1–7.

      [11] Mohamed NS, Mamat M & Rivaie M (2017) A new nonlinear conjugate gradient coefficient under strong Wolfe-Powell line search. AIP Conference Proceedings 1870, 1–7.

      [12] Hamoda M, Mamat M, Rivaie M, Salleh Z & Amani Z (2015), A new nonlinear conjugate gradient coefficient for unconstrained optimization. Applied Mathematical Sciences 9, 1813–1822.

      [13] Abashar A, Mamat M, Rivaie M, Mohd I & Omer O (2014), The proof of sufficient descent condition for a new type of conjugate gradient methods. AIP Conference Proceedings 1602, 293–303.

      [14] Ghani NHA, Rivaie M & Mamat M (2016), A modified form of conjugate gradient method for unconstrained optimization problems. AIP Conference Proceedings 1739, 1–8.

      [15] Hajar N, Mamat M, Rivaie M & Jusoh I (2016) A new type of descent conjugate gradient method with exact line search. AIP Conference Proceedings 1739, 1–8.

      [16] Zoutendijk G (1970), Nonlinear programming computational methods. Integer and Nonlinear Programming 143, 37–86.

      [17] Shoid S, Rivaie M & Mamat M (2016), A modification of classical conjugate gradient method using strong Wolfe line search. AIP Conference Proceedings 1739, 1–8.

      [18] Shapiee N, Rivaie M & Mamat M (2016), A new classical conjugate gradient coefficient with exact line search. AIP Conference Proceedings 1739, 1–8.

      [19] Abidin ZZ, Mamat M, Rivaie M & Mohd I (2014), A new steepest descent method. AIP Conference Proceedings 1602, 273–278.

      [20] Buonaccorsi JP & Elkinton JS (2002), Regression analysis in a spatial-temporal context: Least square, generalized least square and the use of the bootstrap. Journal of Agricultural, Biological and Environment Statistics 7, 4–20.

      [21] Department of Statistics Malaysia (DOSM) (2017), LFS, Unemployment rate in Malaysia from 1982-2014. DOSM.

  • Downloads

  • How to Cite

    Syarafina Mohamed, N., Mamat, M., Rivaie, M., Hamizah Abdul Ghani, N., Zull, N., & Syoid, S. (2018). Estimating the unemployment rate using least square and conjugate gradient methods. International Journal of Engineering & Technology, 7(2.15), 94-97. https://doi.org/10.14419/ijet.v7i2.15.11360

    Received date: 2018-04-10

    Accepted date: 2018-04-10

    Published date: 2018-04-06