Want to apply AI/ML in your lab experiments but don’t know where to start?

The first step is simpler than you might think. 

Here are some practical tips:

1) Start with the background

Before running experiments, review scientific literature. In addition to understanding the fundamentals, note down the variables and responses reported by the authors (including ranges and units). Record everything that’s relevant,

2) Think in terms of “inputs” and “outputs”

Inputs: Variables you’ll control in the lab (e.g., temperature, pH, time, concentrations, type of substrate). Include all potential candidates—don’t discard anything at the start.

Outputs: Results measured after the experiment (e.g., yield, mg/L, purity, growth). If any are qualitative, register it too,

3) Structure your data in a spreadsheet

row A: Trial ID | Date | Input variables (X1, X2, …) | Output variables (Y1, Y2, …)

4) Consistency above all

Use clear names, fixed units, and stable formats (dates, decimals). Avoid empty cells; if data is missing, mark it as NA. This makes model training much easier.

5) Capture the context

Beyond the results, record conditions and equipment used (pH, temp, mg/L, instrument, protocol version). This traceability helps ML detect real patterns.

Getting started isn't difficult: it's a matter of order and perseverance. If you have any questions, please ask me. I'll be happy to help.

#MachineLearningInScience #DataDrivenResearch 

#ExperimentalDesign