Topic 2:- Machine Learning Workflow

Workflow 

The Machine Learning (ML) workflow is a step-by-step process used to build, train, test, and evaluate an ML model.

It ensures that data is properly prepared, the model is correctly trained, and its performance is accurately measured.

 Main Stages of the ML Workflow

ML Overview(step-by-step)

1. Data Collection

• Gather data from various sources such as CSV files, databases, APIs, sensors, or online datasets.
• Example: Collecting house price data (area, location, price).

2. Data Preprocessing

• Clean and prepare data before training.
• Handle missing values, outliers, and categorical encoding.
• Apply feature scaling and normalization.

3. Train/Test Split

Dataset Division

• Split the dataset into two parts:
• Training Set: Used to train the model (70–80% of data)
• Testing Set: Used to test the model (20–30% of data)

 Purpose:
To check how well the model performs on unseen data (generalization).

 4.  Model Training

• Choose an algorithm (e.g., Linear Regression, Decision Tree).
• Feed the training data to the model to help it learn patterns.

5.  Model Testing / Prediction

• Use the test data to check model predictions.

6. Evaluation Metrics

Used to measure how good your model is.
Metrics depend on the type of ML problem.

 

For Regression Problems:

Metric	Description	Function
MAE (Mean Absolute Error)	Average of absolute errors	mean_absolute_error()
MSE (Mean Squared Error)	Average of squared errors	mean_squared_error()
R² Score	Accuracy of regression fit	r2_score()

 

 For Classification Problems:

Metric	Description	Function
Accuracy	Percentage of correct predictions	accuracy_score()
Precision	True positives among predicted positives	precision_score()
Recall	True positives among actual positives	recall_score()
F1-Score	Harmonic mean of Precision & Recall	f1_score()
Confusion Matrix	Table showing TP, FP, TN, FN	confusion_matrix()

 7.  Model Optimization

• Tune hyperparameters (e.g., learning rate, max_depth).
• Use Grid Search or Random Search for tuning.
• Perform Cross-Validation for better evaluation.

 8.  Deployment

• Deploy your trained model using Flask, Streamlit, or FastAPI for real-world use.

 Summary

The Machine Learning workflow is a structured pipeline involving data collection, preprocessing, train/test splitting, model training, evaluation, and deployment ensuring reliable and accurate model performance.