DEMAND FORECASTING FOR LESS LETHAL AMMUNITION AND GRENADES IN THE MILITARY POLICE
Name: THIAGO PAVAN SILVA
Publication date: 28/05/2020
Advisor:
Name |
Role |
|---|---|
| HÉLIO ZANQUETTO FILHO | Advisor * |
Examining board:
| Name |
Role |
|---|---|
| ADONAI JOSÉ LACRUZ | Internal Examiner * |
| HÉLIO ZANQUETTO FILHO | Advisor * |
| MARCOS PAULO VALADARES DE OLIVEIRA | Internal Examiner * |
Summary: The research aimed to develop a demand forecast model for less-lethal ammunition and grenades in the military police sphere. To this end, a case study was conducted at the Military Police of the State of Espírito Santo, aiming to identify causal factors that influence the demand for less-lethal ammunition and grenades, assess the predictive capacity of the forecasting model and analyze the relevance of the causal factors that influence demand for less-lethal materials. A mixed-methods approach was adopted, implementing the sequential strategy recommended by Creswell (2010), which in the qualitative phase sought to collect the variables that affect the demand for less lethal materials through semi-structured interviews with managers of war materials and in order requisition documents. Thus, it was initially possible to construct the causal variables that affect the demand for less lethal materials, following the content analysis from the perspective of Bardin (2011). Finally, an exploratory factor analysis was carried out, allowing a more parsimonious forecasting model with ten causal factors. In the quantitative phase of the research, the predictive capacity of the factors was evaluated, using the Random Forest forecasting method, with cross-validation of 5-fold and 10-fold for the products of pepper spray, controlled impact ammunition and moral effect grenade. Thus, the research makes it possible to understand, from the perspective of military police officers, the factors that influence the demand for less lethal ammunition and grenades, in addition to presenting a model to support the decisions of military managers when planning demand through the Random Forest technique.
