MLFriend: Interactive Prediction Task Recommendation for Event-Driven Time-Series Data
This addresses the problem of reducing reliance on human experts for task definition in machine learning, particularly for event-driven time-series data, representing an incremental advance in automation.
The paper tackles the challenge of automatically defining useful prediction tasks on event-driven time-series data, introducing MLFriend, which generates and recommends tasks, achieving 722 useful tasks out of 2885 generated and identifying top 10 tasks within a batch of 100.
Most automation in machine learning focuses on model selection and hyper parameter tuning, and many overlook the challenge of automatically defining predictive tasks. We still heavily rely on human experts to define prediction tasks, and generate labels by aggregating raw data. In this paper, we tackle the challenge of defining useful prediction problems on event-driven time-series data. We introduce MLFriend to address this challenge. MLFriend first generates all possible prediction tasks under a predefined space, then interacts with a data scientist to learn the context of the data and recommend good prediction tasks from all the tasks in the space. We evaluate our system on three different datasets and generate a total of 2885 prediction tasks and solve them. Out of these 722 were deemed useful by expert data scientists. We also show that an automatic prediction task discovery system is able to identify top 10 tasks that a user may like within a batch of 100 tasks.