Please note this is not a comprehensive article of PCA. I’m just sharing some key thoughts on PCA that might be helpful to others who are confused by this technique. The basic idea is that not every feature is necessarily useful/important. Taking the Iris dataset for example. Original dataset has…

The first time I learned about neural networks was November 2019 when taking the famous machine learning course by Andrew Ng. I had a hard time understanding the algorithm (not gonna lie). Early this year I spent some time studying neural networks again by referring many other resources. …

I would like to mark down the reasons that I found helpful for intuitive understandings. First reason: for the sake of familiarity. Distributions whose variances stay the same across all the data is the one we are most familiar with and we are happy to deal with! But some variables…

A good example to intuitively understand the concept: Say you have 100 dollars to buy 5 things. You can buy whatever you want for the first 4 things as long as there’s money left. But for the fifth thing, the amount of money you can spend is restricted at the moment you bought the fourth thing. You cannot choose whatever you want anymore. In this case, the degree of freedom is 4, because you can freely choose the first 4 things.

I first came across t-test when I was studying quantitaive research on communication science back in college. Recently I did some research about t-test as well as t-distribution, hopefully to get a deeper understanding. I would like to memo down what I have figured out. the purpose of t-test is…

When pre-processing training data and test data, does it make a difference to pre-process before splitting or to split before pre-processing? The principle here is to make sure that the model does not gain any sights from test data. …

Revised on 4/21 I would like to share what I’ve come to understand. ・First, I would like correct my misconception that Xgboost does not add up trees. In fact, it does add trees. ・To add a tree, we need to set a learning rate ( called “eta” in the xgboost…

I’ve tried many ways to pip install but kept running into the following error: XGBoostLibraryNotFound: Cannot find XGBoost Library in the candidate path, did you install compilers and run build.sh in root path? Eventually I gave up pip and this worked for me: In terminal run: $ git clone — recursive https://github.com/dmlc/xgboost.git This should downloads a folder called “xgboost”. Go to /xgboost/python-package

In week 2, after I downloaded the exercise zip file “machine-learning-ex”, I found that I could not unzip it in Mathlab online. Here is the fix: After clicking the machine-learning-ex provided in the session, when the window pop up, DO NOT click on SAVE but choose OPEN WITH, then click OK. You will get a zip local file named as “machine-learning-ex-1.zip”. Don’t be mistaken by the file name. It actually includes all the exercise.