The ML solution development process
ML offers many possibilities to augment and automate business. To get the best from ML, teams and people engaged in ML-driven business transformation need to understand both ML and the business itself. Efficient business transformation begins with having a rough understanding of the business, including aspects such as value-chain analysis, use-case identification, data mapping, and business simulations to validate the business transformation. Figure 2.1 presents a process to develop ML solutions to augment or automate business operations:
Figure 2.1 – ML solution development process
Business understanding is the genesis of developing an ML solution. After having a decent business understanding, we proceed to data analysis, where the right data is acquired, versioned, and stored. Data is consumed for ML modeling using data pipelines where feature engineering is done to get the right features to train the model. We evaluate...
