PredictScript is a high-level API for Prediction-as-a-Service.
There are two API calls:
predict(modelName, features)- computes the probability an event will occur and returns that probability.
label(prediction.id, outcome)- reports what happened (true / false). PredictScript uses labels to train the model.
In PredictScript a model is a name and the associated predictions and labels you've sent.
If you are familiar with ML you may be wondering "How do I train my model?" The answer is that
label() do the training for you. No need to build a training set or tune parameters. Once models have enough data they'll start serving predictions automatically.
This page will walk you through getting up and running.
First sign up for an account. It's free and you'll get an API key that you can use to start training models.
Add these two lines to the end of <body> on pages where you'll be making a prediction:
<script src="https://predictscript.com/assets/predictscript.js"></script> <script>predictscript.config('API_KEY_HERE')</script>