You have probably heard of the Swiss army knife. If not, just take a look at the image below. It contains many blades and tools. Each one is specialized at a particular task. In some cases, different blades can do the same task but with a different degree of performance.

I think of the machine learning algorithms as Swiss army knife. There are many different algorithms. Certain tasks require to use a particular algorithm whereas some tasks can be done with many different algorithms. The performance might change depending on the characteristics of the task and data.

In this post, I will share 16 tips that I think will help you better understand the algorithms. My goal is not to explain how algorithms work in detail. I will rather give some tips or details about them.

Some tips will be more general and not be focused on a particular algorithm. For instance, log loss is a cost function that is related to all classification algorithm.

I will assume you have a basic understanding of the algorithm. Even if you don’t, you can pick some details that will help you later on.

Let’s start.

1. C parameter of support vector machine (SVM)

C parameter of SVM adds a penalty for each misclassified data point. If c is small, penalty for a misclassified point is low so a decision boundary with a large margin is chosen at the expense of a greater number of misclassifications.

If c is large, SVM tries to minimize the number of misclassified examples due to high penalty which results in a decision boundary with a smaller margin. Penalty is not same for all misclassified examples. It is directly proportional to the distance to decision boundary.

#artificial-intelligence #machine-learning #data-science #data-analysis