Dimensionality Reduction (PCA) | Machine Learning

Dimensionality Reduction (PCA) | Machine Learning



This is about Dimensionality Reduction which is considered very important nowadays. If we see that the data is huge, we can have multiple columns or classes. There are not just two dimensions but the data is divided into multiple dimensions. As discussed earlier, the more the dimensions, there is more complexity and uncertainty and the chances of errors are huge. More errors indicate lesser accurate predictions.

There are methods for reducing dimensions in statistics and those same methods used in machine learning are also known as Principal Component Analysis (PCA). It helps us in reducing the dimensions and produce a new dataset with lesser number of dimensions. But it is important to mention that PCA results in loss of information because the dimensions are reduced and we cannot expect 100% correct predictions. We cannot call it a mathematical model as it is not a prediction technique. But we use it in prediction technique for classification algorithms and also use it to reduce the number of dimensions so that the time complexity is reduced and we get better predictions.

Join Our Data Science Bootcamp

With Job Assistance

+91 892 908 3443

what's app