To build a machine-learning model, we normally follow a step-by-step process. We can divide all the steps into three groups. Those are:
- Data Pre-Processing
After collecting the data, the first thing we have to do is clean the data. Data cleaning is the process of fixing or removing incorrect, corrupted, incorrectly formatted, duplicate, or incomplete data within a dataset. After cleaning the data, we have to split the data into training & test sets. We do it to avoid overfitting. I will explain it more in the upcoming blogs. Now we have to do feature scaling. This helps to ensure that features with different units of measurement or scales do not dominate the model training process.
Now comes the most interesting part, model building. We have to build a model and train it using the training set that we got after splitting our data. We have to decide which model we should use according to the data we have. After completing the training part, now our model is ready to make predictions.
And finally, we move on to evaluations. We will calculate some performance metrics and make a verdict about our model, whether it's a good-fitting model and if it works for our data or not. And this is a very important step to make sure that the models we build really serve the purpose that they're designed for.
Latest comments (0)