ADS Capstone Chronicles Revised

‭17‬

‭Figure 4.10‬ ‭Accidents by Hour of Day and Weather Condition‬

‭Figure‬ ‭4.11‬ ‭presented‬ ‭a‬ ‭correlation‬ ‭heatmap‬ ‭for‬ ‭all‬ ‭of‬ ‭the‬ ‭numerical‬ ‭variables‬‭in‬‭the‬‭dataset.‬‭The‬ ‭heatmap‬ ‭visually‬ ‭represented‬ ‭the‬ ‭relationships‬ ‭between‬ ‭the‬ ‭variables,‬ ‭with‬ ‭darker‬ ‭colors‬ ‭indicating‬ ‭stronger‬ ‭correlations.‬ ‭Several‬ ‭pairs‬ ‭of‬ ‭variables‬ ‭showed‬ ‭high‬ ‭correlation‬ ‭with‬ ‭one‬ ‭another,‬ ‭suggesting‬ ‭potential‬ ‭redundancy‬ ‭in‬ ‭the‬ ‭dataset.‬‭Upon‬‭further‬‭inspection,‬‭it‬‭was‬‭noted‬‭that‬ ‭these‬ ‭highly‬ ‭correlated‬ ‭variables‬‭were‬‭duplicates‬ ‭or‬ ‭highly‬ ‭similar‬ ‭measures‬ ‭of‬ ‭the‬ ‭same‬ ‭underlying‬ ‭concept.‬ ‭These‬ ‭duplicated‬ ‭variables‬ ‭were‬ ‭addressed‬ ‭during‬ ‭the‬ ‭feature‬ ‭selection‬ ‭and‬ ‭engineering‬‭phase‬‭of‬‭the‬‭project.‬‭By‬‭removing‬‭or‬ ‭consolidating‬ ‭these‬ ‭redundant‬ ‭features,‬ ‭the‬ ‭efficiency‬ ‭of‬ ‭the‬ ‭model‬ ‭was‬ ‭improved,‬ ‭and‬ ‭the‬ ‭risk‬ ‭of‬ ‭multicollinearity,‬ ‭which‬ ‭could‬ ‭negatively‬ ‭affect‬ ‭the‬ ‭performance‬‭and‬‭interpretability‬‭of‬‭the‬ ‭model, was reduced.‬

257