MLOps framework

The below are the comprehensive steps must needed for a model training on any dataset.

Use the model_params file under config section for any hyper param definitions

Data

Data preparation Data preprocessing Data cleansing Data wrangling

Data Encoding Data scaling Data normalization

Visualize Data

Feature Engineering Clustering Dimensionality Reduction

Visualize Data

k-fold define CrossValidation Preparation

Model

TrainModel |- DecisionTree |- Boosting |- SVM (SUPPORT VECTOR MACHINES) |- KNN (K-NEAREST NEIGHBOURS) |- NN (NEURAL NETWORKS)

Score Logging |- Accuracy |- Precision |- Recall |- F1Score

Visualize Score

Score visualization |- Train score plot |- Validation score plot

Score analysis |- Bias Analysis |- Variance Analysis |- Overfitting Analysis |- Underfitting Analysis

Visualize time

Time analysis |- Training time |- Validation time |- Epoch Time

Save the model checkpoints Version the models

Retrain the models

** Repeat

Select the Model Version the model Save the checkpoint

########

Deploy to Production