Google Play Expected Downloads Calculator

• 

  For those trying to get the money

Inspiration

If you are a prospective app developer, it would be good to know what your app's expected number of downloads would be.

What it does

Takes in your app's category, price, size in MB, content rating, and genre to predict how many downloads your app is expected to receive.

How we built it

We used python and scikit-learn to implement a k-nearest-neighbors predictive model. Then we used jswing to make a more intuitive GUI for the user to input their information.

Challenges we ran into

We were unable to run our python script within our java application. Instead, we had to run the java application and the python file separately, rather than have them neatly integrated into a single application.

Accomplishments that we're proud of

We are proud of completing the model and being able to apply the model to predict individual samples.

What we learned

It's hard to run python in java. Data needs to be cleaned extensively before a good model is made.

What's next for Play Store Expected Downloads Calculator

We will compile everything into a single, easy to use GUI or website rather than use the clunky command prompt. We will also try to incorporate a model for Apple Store apps and as with all machine learning techniques, work on improving our model. We also want our applet to return the top 3 most likely outcomes, rather than just a single outcome to give the user a better idea of what to expect.

Built With

• anaconda
• java
• jswing
• python
• scikit-learn