Week 2 — Career Path Predictor

Hello everyone, and welcome to our second week blog post of our Career Path Predictor Project. Last week, we left behind a week full of Artificial Intelligence as we wished for you. Let’s take a look at the progress in our project together.

We started this week by creating a Github Repository so we can access and see our progress with our tutorial. You can reach our repository by clicking HERE, and you can see the final version of our project and our commits at the end of this week’s work. In the meantime, let’s summarize our work for you while you look at the link..

Reading & Understanding Data:

After setting the workspace, we started reading and understanding our data.

With the images we have added above, you can have the necessary information about our dataset. As can be seen from the images, there is no NaN value in our dataset, only 4 features in 19 different columns except the label column are given numerically. As you will notice, we were disturbed by the irregular appearance of the column names, and we arranged the column names first to make them more understandable for everyone. Then we wanted to have information about the outliers in our data, but after performing the checks, we realized that there was no outlier in our data.

Starting to Preprocessing Steps:

We have made our categorical columns, which are too many, numeric with the label encoder. Thus, we were able to use these data for PCA analysis and to see correlations in the remaining steps. As a result of PCA analysis, we detected a column that could be dropped, but we decided not to drop it considering that the positive effects outweigh the negative effects.

We checked the label column for any imbalances to see if any sampling was needed as a final operation. We were happy with the results, because we saw that there was no big imbalance in the data. We would like to thank the people who created the dataset.

We are still keeping our excitement from last week. We hope you still share our excitement. Thank you for staying tuned. See you next week.

• Melike Nur Dulkadir
• Sare Naz Ersoy