Identify the specific business problem to solve. Establish clear business objectives and success criteria to evaluate the model. Align stakeholders to gain consensus on goals and define the target.
ML Pipeline:
A series of steps starting with a business goal and ending with a deployed ML model.
Includes defining the problem, collecting and preparing data, training the model, deploying, and monitoring.
It’s an iterative process, repeated even after deployment.
Business Goal:
Start by identifying the business goal and measuring success using clear criteria.
Ensure the problem can be solved with ML.
Data:
Ensure relevant and high-quality data is available for model training.
Evaluate data sources for accessibility and correctness.
ML Approach:
Start with the simplest solution and perform a cost-benefit analysis to decide if ML is the right choice.
AWS provides pre-trained models and AI services to save time and cost.
AWS Services:
AWS offers AI services like Amazon Comprehend, Amazon SageMaker, and Amazon Bedrock to help build models.
These services allow you to customize pre-trained models or fine-tune them with your own data.
Building Models:
The hardest and most costly option is building a custom model from scratch.
