• 

  This graph shows the learning process

• 

  This sheet shows the sample prediction result with an accuracy of 0.85

Rice-Datathon

Using the Perceptron Learning Model to predict divorce based on answers from a questionnaire

We picked the divorce issue as our challenge because we would like to apply what we've learned in CS into areas other than STEM.

We don't have much experience with machine learning, so we studied the popular algorithms first. We decided that perceptron learning is a good fit for this problem, as the outcome is either 0(divorce) or 1(no divorce)

We faced some challenges playing around with pandas and numpy modules. We used to write in codeskulptor in Comp140, so writing using desktop editors is a brand new thing for us. Fortunately we made it!

Thanks for this opportunity!

Built With

• python