Anomaly detection-model selection

Hi,
I am a beginner of Machine Learning. These days, I tried a anomaly detection project by myself. I am confused on model selection. Since after I clean my main data(In one table I have 4 columns data all of them need find the outlier - > features = 4.). However, when I tried to fit the data with the density estimation model I find I can’t. Since after I plot a scatter of them, I find each of them shows a linear relationship, and they have very high correlation. It’s not contain cluster, just linear with some very obvious outliers.

Dose anyone can provide me some suggestions which model should I need to use on this symptom? Since I only know how to use density estimation. I want to try classification before, but it’s unsupervised(I only have those four columns, I think it may unsupervised, but I am not sure). I finished the ML course yesterday, so I am not very familiar with all of them, I am not sure. Dose any one can help me? Thank you so much~

language: python
Unsupervised Learning, Recommenders, Reinforcement
Supervised ML: Regression and Classification
AI Discussions

What are the axes on your plot? Is that plot showing just one of your four features on the horizontal axis and the output value on the vertical axis?

Also, what Machine Learning courses have you attended?