kaggle

A collection of 7 posts

Predicting Animal Shelter Outcomes

Predicting Animal Shelter Outcomes

I was preparing to give a talk for some important Veteran Advocacy Non-Profits here in DC the other day, and wanted to provide an easy to understand example of machine learning and classification.

Amazon Food Reviews with Keras

Amazon Food Reviews with Keras

I'm teaching myself TensorFlow and Keras. The entire principle of using data flow graphs and deep learning is really fascinating, but it's a learning curve for sure. I've been poking around the MNIST

Amazon Food Reviews, Part V

Amazon Food Reviews, Part V

In this final task, my goal is to predict the Amazon score (1 - 5) based on the reviews - a multiclass text classification problem. This post discusses the use of five pipeline

Amazon Food Reviews, Part IV

Amazon Food Reviews, Part IV

In Part I we did some exploratory analysis after using the TextBlob package to apply sentiment scores using its polarity method. We found the mean sentiment was around .24 and then spilt up

Amazon Food Reviews, Part III

Amazon Food Reviews, Part III

Now that we know a little about the extremes of perfectly positive and negative sentiment reviews, let's analyze some key features of the text. Here we'll allow a user to retrieve the top

Amazon Food Reviews, Part II

Amazon Food Reviews, Part II

Ok, so now that we explored the contents of the Amazon Fine Reviews data file, it's time to move on to the second task: How many reviews and products are perfectly positive and