From the course: Certified Analytics Professional (CAP) Cert Prep

Unlock the full course today

Join today to access over 25,000 courses taught by industry experts.

Tracking model quality

Tracking model quality

From the course: Certified Analytics Professional (CAP) Cert Prep

Start my 1-month free trial Buy for my team

Tracking model quality

“

- [Instructor] Model quality can deteriorate due to various factors. For one, your original assumptions about business goals may not be valued anymore. Maybe your model no longer offers sufficient value due to changes in business priorities. Quality of your data could change too affecting the performance of the model. Because of this possibility of declining model quality tracking with specific evaluation criteria is imperative. The most obvious criterion is accuracy. We already discussed how to measure it. Whether your model provides new insights or not is another important criteria. Consistency is also critical if model behaves erratically due to expected variations in data on investigation and its integrity is necessary. Finally, overfeeding is what you should look out for. It's an established practice to train a model using a portion of a data set. Overfitting occurs when you calibrate the model too much based on…

Contents