Understanding regression in supervised machine learning
Regressions are models generally used to determine the relationship or correlation between dependent and independent variables. Within the context of machine learning, we define regressions as supervised machine learning models that allow for the identification of correlations between two or more variables in order to generalize or learn from historical data to make predictions on new observations.
Within the confines of the biotechnology space, we use regression models to predict values in many different areas.
- Predicting the LCAP of a compound ahead of time
- Predicting titer results further upstream
- Predicting the isoelectric point of a monoclonal antibody
- Predicting the decomposition percentages of compounds
Correlations are generally established between two columns. Two columns within a dataset are said to have a strong correlation when a dependence is observed. The specific relationship can be...
