Model approval workflows in MLOps - Mini Project: Build & Apply

Model Approval Workflows

📖 Scenario: You work in a team that builds machine learning models. Before a model is used in production, it must be approved based on its performance metrics. This project helps you create a simple approval workflow to decide if a model passes or needs review.

🎯 Goal: Build a small program that stores model performance scores, sets an approval threshold, checks which models meet the threshold, and prints the approved models.

📋 What You'll Learn

Create a dictionary called models with model names as keys and their accuracy scores as values

Create a variable called approval_threshold with a float value representing the minimum accuracy needed for approval

Use a dictionary comprehension to create a new dictionary approved_models containing only models with accuracy greater than or equal to approval_threshold

Print the approved_models dictionary

💡 Why This Matters

🌍 Real World

Teams building machine learning models need to approve models before using them in real applications to ensure quality and reliability.

💼 Career

Understanding model approval workflows is important for roles in MLOps, data science, and machine learning engineering to maintain production standards.

1

Create the model performance data

Create a dictionary called models with these exact entries: 'model_A': 0.82, 'model_B': 0.76, 'model_C': 0.91, 'model_D': 0.68

MLOps

# Create the models dictionary with exact accuracy scores
# Your code here

Need a hint?

Use curly braces {} to create a dictionary with keys as model names and values as accuracy scores.

2

Set the approval threshold

Create a variable called approval_threshold and set it to 0.80 to represent the minimum accuracy needed for model approval

MLOps

models = {'model_A': 0.82, 'model_B': 0.76, 'model_C': 0.91, 'model_D': 0.68}
# Set the approval_threshold variable to 0.80
# Your code here

Need a hint?

Use a simple assignment to create the approval_threshold variable.

3

Filter approved models

Use a dictionary comprehension to create a new dictionary called approved_models that contains only the models from models with accuracy greater than or equal to approval_threshold

MLOps

models = {'model_A': 0.82, 'model_B': 0.76, 'model_C': 0.91, 'model_D': 0.68}
approval_threshold = 0.80
# Create approved_models dictionary using dictionary comprehension
# Your code here

Need a hint?

Use {name: acc for name, acc in models.items() if acc >= approval_threshold} to filter the dictionary.

4

Display approved models

Write a print statement to display the approved_models dictionary

MLOps

models = {'model_A': 0.82, 'model_B': 0.76, 'model_C': 0.91, 'model_D': 0.68}
approval_threshold = 0.80
approved_models = {name: acc for name, acc in models.items() if acc >= approval_threshold}
# Print the approved_models dictionary
# Your code here

Need a hint?

Use print(approved_models) to show the filtered dictionary.