ATT PROGNOSTISERA DET FINANSIELLA RESULTATET I SVENSKA KOMMUNER EN JÄMFÖRELSE MELLAN ETT ELASTISKT NÄT OCH EN BASELINE
(2020) STAH11 20192Department of Statistics
- Abstract
- Machine learning techniques have gained ground in macroeconomic forecasting in recent years. Estimations with the elastic net technique have yielded good results. There is currently not much research on how statistical techniques can be used to estimate the financial result in Swedish municipalities. This study adds to the current body of research by investigating how the elastic net technique can be used to forecast the financial result in Swedish municipalities.
Data is collected from several sources and is used to forecast the financial result in Swedish municipalities. Estimations are carried out using an elastic net and a baseline in the form of OLS and are then compared and discussed. The results show that the estimations using... (More) - Machine learning techniques have gained ground in macroeconomic forecasting in recent years. Estimations with the elastic net technique have yielded good results. There is currently not much research on how statistical techniques can be used to estimate the financial result in Swedish municipalities. This study adds to the current body of research by investigating how the elastic net technique can be used to forecast the financial result in Swedish municipalities.
Data is collected from several sources and is used to forecast the financial result in Swedish municipalities. Estimations are carried out using an elastic net and a baseline in the form of OLS and are then compared and discussed. The results show that the estimations using an elastic net give more accurate forecasts than OLS in Swedish municipalities between 2002 and 2018.
The most accurate forecasting method in this study was the Ridge regression. The study concludes by suggesting how the financial result in Swedish municipalities could be more thoroughly investigated using statistical methods. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9033365
- author
- Nilsson, Sofia LU
- supervisor
- organization
- course
- STAH11 20192
- year
- 2020
- type
- M2 - Bachelor Degree
- subject
- keywords
- Elastic net, Ridge regression, Lasso regression, OLS, finances of Swedish municipalities, prediction, forecasting
- language
- Swedish
- id
- 9033365
- date added to LUP
- 2021-01-07 12:18:23
- date last changed
- 2021-01-15 03:41:04
@misc{9033365, abstract = {{Machine learning techniques have gained ground in macroeconomic forecasting in recent years. Estimations with the elastic net technique have yielded good results. There is currently not much research on how statistical techniques can be used to estimate the financial result in Swedish municipalities. This study adds to the current body of research by investigating how the elastic net technique can be used to forecast the financial result in Swedish municipalities. Data is collected from several sources and is used to forecast the financial result in Swedish municipalities. Estimations are carried out using an elastic net and a baseline in the form of OLS and are then compared and discussed. The results show that the estimations using an elastic net give more accurate forecasts than OLS in Swedish municipalities between 2002 and 2018. The most accurate forecasting method in this study was the Ridge regression. The study concludes by suggesting how the financial result in Swedish municipalities could be more thoroughly investigated using statistical methods.}}, author = {{Nilsson, Sofia}}, language = {{swe}}, note = {{Student Paper}}, title = {{ATT PROGNOSTISERA DET FINANSIELLA RESULTATET I SVENSKA KOMMUNER EN JÄMFÖRELSE MELLAN ETT ELASTISKT NÄT OCH EN BASELINE}}, year = {{2020}}, }