Situations and Challenges: Full effort required for classification

You belong to the quality control department of a company that makes PC peripherals. The new product released 3 months ago has had many breakdowns and repairs. At each site, information on malfunctions is reformatted into a common format for classification. You would like to continue to do this in order to understand the situation, but one site told you that they would like to stop classifying the types of failures because they are too difficult and require a lot of man-hours.

Using predictive analytics: Replace classification with predictive analytics

• Prepare failure information data (you will use sample datasets in the tutorial)
• Use Easy Predictive Analytics to determine the type of the failure information
• Fewer man-hours by entrusting type determination of failure information to Easy Predictive Analytics

Expected Effect: Reduction of effort. Stable labeling results.

Automation of failure labeling reduces man-hours. When the person in charge changes, the contents of the labeling may also change, but automation is expected to make it more uniform.

Now let’s actually use Easy Predictive Analytics to categorize failure types based on predictions!