AutoML & HPO for Omics Pipelines — Hands-on

Learn how to design and operate AutoML workflows for omics and clinical machine learning. This module covers search space design, hyperparameter optimization algorithms, configuration management, experiment tracking and best practices for robust, reproducible AutoML pipelines in R and Python.

AutoML & HPO for Omics Pipelines

Help Desk · WhatsApp

Session Index

Session 1 — AutoML Concepts & Search Spaces Session 2 — Hyperparameter Optimization Algorithms Session 3 — AutoML Pipelines for Omics & Clinical Data Session 4 — Benchmarking, Governance & Deliverables

Session 1

Fee: Rs 8800

AutoML Concepts & Search Spaces

What AutoML solves in bioinformatics workflows

automation vs manual tuning typical omics and clinical use cases tradeoffs: speed, accuracy, interpretability

Defining configuration and search spaces

hyperparameters vs fixed design choices continuous, integer and categorical spaces constraints and conditional parameters

Cost, budget and resource considerations

time and compute budget planning parallel vs sequential evaluations early stopping and pruning ideas

Session 2

Fee: Rs 11800

Hyperparameter Optimization Algorithms

Search strategies and baselines

grid search and random search Latin hypercube and low discrepancy ideas when simple methods are enough

Bayesian and adaptive optimization

Bayesian optimization intuition Tree Parzen Estimators style approaches successive halving and Hyperband logic

Practical tuning workflows for omics models

optimizing pipelines, not just models multi metric and constrained optimization logging and reproducing best runs

Session 3

Fee: Rs 14800

AutoML Pipelines for Omics & Clinical Data

Building end to end search pipelines

preprocessing and modeling steps together feature selection inside pipelines stratified cross validation integration

Handling high dimensional omics features

filtering and dimensionality reduction options sparse models and regularization paths small n, large p constraints

Integrating clinical and omics covariates

heterogeneous feature handling group wise preprocessing strategies balanced optimization objectives

Session 4

Fee: Rs 18800

Benchmarking, Governance & Deliverables

Benchmark design and fair comparisons

fixed data splits and seeds reporting central tendency and spread avoiding optimistic bias in selection

Governance, logging and experiment tracking

tracking runs, configs and metrics replay and rollback of experiments alignment with MLOps practices

Deliverables: tuned pipeline and report pack

final AutoML configuration and code comparison tables and plots written summary for methods section

PDF

AutoML & HPO for Omics Pipelines Training | Biostatistics & ML for Omics

AutoML & HPO for Omics Pipelines — Hands-on