Abstract
In multiple regression analysis, we aim to construct a statistical model that describes the relationship between a dependent variable and several independent variables. However, the data may contain observations that differ significantly from the rest of the values. These observations are known as \"outliers.\"
Outliers are data points that fall outside the general pattern of relationships between variables. They may result from measurement or input errors, or they may reflect exceptional cases with real significance. The presence of these outliers can distort the results of the analysis.
Hence, the aim of the research was to build an algorithm to detect the strays present in the data and then delete them from the data to reach the most accurate results. The algorithm was applied to data free of outliers and data containing 10%, 20% and 30% of outliers cases. The algorithm proved its efficiency in all cases.
Outliers are data points that fall outside the general pattern of relationships between variables. They may result from measurement or input errors, or they may reflect exceptional cases with real significance. The presence of these outliers can distort the results of the analysis.
Hence, the aim of the research was to build an algorithm to detect the strays present in the data and then delete them from the data to reach the most accurate results. The algorithm was applied to data free of outliers and data containing 10%, 20% and 30% of outliers cases. The algorithm proved its efficiency in all cases.