Demystifying Machine Learning

Summary:

• The regression task aims to predict continuous numbers.
• Simple regression outputs only one continuous number.
• The aim is to fit a model so that it’s as close as possible to the training examples.
• The model can be parametric (linear, quadratic, cubic, etc.) or non-parametric.
• Assuming a linear model, the model is a line (one input feature), a plane (two input features), or a hyperplane (more than two features).
• Fitting a model means learning the line/plane/hyperplane that ‘best’ fits the data.