WK 3 – DATA 610 6/3/20 –

Since both method are preferable for different data scenarios. There are cases where Davenport will work better than CRISP-DM and vice versa. For Davenport, I love its simplicity and a more traditional outlook. For example, the model design comes before data collection, and thus it allows one to have a wide range of choices when it comes to modeling. Although it required to collect and prepare data before modeling in CRISP-DM, it also allows one to move back and forth while making changes until correct data and models are made available. Clearly, CRISP-DM control how ones deal with data and modeling and thus may likely reduce errors in data and model. Besides, CRISP-DM is the most famous and widely used methodology. Therefore, it has a wide range of content and guides for learning. Based on these aspects, I prefer CRISP-DM.