PCA in Everyday Life

What is PCA in Machine Learning?

PCA stands for Principal Component Analysis. It is a way of finding the most important features in a dataset. For example, if you have a dataset of pictures of dogs, PCA could find the features that make a dog look like a dog, such as its shape, size, and color.

Correlation:

For me, PCA is about school days to college days and all the subjects we studied in our life span, but if we see today we are merging all the subjects which correlates us today with our regular daily work of all atleast software engg.

PCA does the same, out of 1000 features in a dataset using PCA we can achieve maximum information from 2 or 3 features after merging it.

So in short , what you are today is the thousand of neurons collected information from your past and learning circumstances, and made you today where you stand right now in akk spheres of life.

What PCA does in Machine Learning?

Principal component analysis is a technique for reducing the number of dimensions in a dataset without incurring a commensurate loss of information. It enjoys a number of uses in machine learning, including visualizing high-dimensional data, anonymizing data, reducing noise, and increasing the ratio of rows to columns by reducing the number of dimensions. It can also be used to perform anomaly detection by measuring the loss incurred when a PCA transform is applied and then inverted. Anomalous samples tend to incur more loss.

Is there any limitation for PCA?

One limitation of using PCA to detect anomalies in multivariate systems is that because it uses linear transforms, PCA is better at modeling linear relationships between variables than nonlinear relationships. Neural networks, by contrast, excel at modeling nonlinear data. That’s the primary reason why state-of-the-art multivariate anomaly detection today commonly relies on deep learning.

Steps how PCA is done on any dataset:

1.Standardize the data.

2. Calculate the covariance matrix

3. Find the eigenvectors and eigenvalues of the covariance matrix

4. Select the number of principal components to keep

5. Project the data onto the principal components.?

Implementation code is available everywhere as per business need, but yeah PCA related to us as human behaviour and we correlate with this model.