Machine Learning vs Statistics
Machine Learning vs Statistics: Historically both disciplines evolved from different perspectives, but with similar
end goals. For example, Machine Learning focused on “prediction” and “decisions”. It relied on “patterns” or “model”
learnt in the process to achieve it. Computation has played key role in its evolution
In contrast, Statistics, founded by statisticians such as Pearson and Fisher, focused on “model learning”. To understand and explain “why” behind a
phenomenon
Probability has played key role in development of the field. As a concrete example, recall the ideal gas law \(PV = nRT\) for Physics. Historically, machine learning only cared about ability to predict \(P\) by knowing \(V\) and \(T\), did not matter how; on the other hand, Statistics did care about the precise form of the relationship between \(P\), \(V\) and \(T\), in particular it being linear. Having said that, in current day and age, both disciplines are getting closer and closer, day-by-day, and this class is such an amalgamation
◀
▶