From the course: Evaluating and Debugging Generative AI
Join today to access over 25,500 courses taught by industry experts.
Solution: Implement bias mitigation
From the course: Evaluating and Debugging Generative AI
Solution: Implement bias mitigation
How did you do? Let me show you how I solved this challenge. Let's start out with the before code. In the before code, we are importing pandas and matplotlib. Your first step is to load the dataset. You'll find the dataset file here underneath the same folder, dataset.csv. You'll load the dataset. Your next step is to analyze the gender distribution. Next, you'll encode the categorical variables. The next step is to separate the majority and minority classes. Make sure you assign those to the df_majority and df_minority variables. You'll use this resample code to upsample the minority class. You'll combine the majority class with the upsampled minority class and display the new class counts. The last step is to train your model on this newly balanced dataset. So let's switch over to see how I solved this problem. The code to load the dataset is pretty straightforward. It's a CSV, so I'm using the pandas read_csv function. Next, I analyze the gender distribution and I plot it. The next…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.