Akansha Yadav clears up a common mix-up: an algorithm is the step-by-step procedure for solving a problem, while a model is the final predictive artifact it produces. She walks through a real ML workflow with concrete examples like handling missing values and selecting features to show how the process becomes the output.
Ladies and gentlemen, hello everyone, welcome back to my YouTube channel. In today's video, we are going to see what is the difference between model and algorithm. Some people do say that this is machine learning model, another says that this is machine learning algorithm. So what do you guys think, that is this different terms or same terms? So this is what we are going to discuss exactly in this video. So let's get back on our screen and see.
So let's see what is model And what is algorithm? If I'm saying algorithm, then it means that I am talking about a way that we are using to tell a story. So the way I am going to tell the story, or the way people are telling stories, that is what algorithm is. If I say in terms of machine learning, the algorithm is a step-by-step process. It means we are defining some steps and we are going to follow those steps to reach on to our predictable outcome. So it's like when we write a computer program, we are going to write the computer program, but in the step-by-step manner we are going to follow that steps, and which in return going to give us the outputs which we have already predicted while defining those steps. So this is what the algorithm is. Another thing that we have is the model. Now what is model? Model is the output. Suppose if we are predicting something and here we reached, like we are having output at this stage. So this output is the model. When everything that we have done is completed and we are here with the output, then that is what we call the model is. So that is what the difference between algorithm and model. So when we are starting our journey, when we are working on the model, like after looking at the data, we define some steps, okay? Okay, this data consists of missing values, so I need to deal with that. So that is also one step which you have mentioned. Another thing you will say that, okay, these are 3 columns which I think can impact to the target column. So in that case, I need to visually look at that data, so I need some graphs. So these steps I'm defining to predict something. So that prediction is the output, which is again called model. And the steps which we have defined is called algorithm.
