The data science loop
A school canteen wants to guess how many idlis to make tomorrow. Where does the work start?
Step 1 of 4
- Collect data. Clean it: fix repeats, wrong values and gaps.
- Train a model on the clean data.
- Check its results with fresh data.
- Train again. The loop repeats.
The work starts with collecting data from past days. Here, that is how many idlis were sold each day.
A model is a program that learns patterns from data. Fresh data is data the model has not seen before.
Data science studies data to find patterns that help people decide.
Cleaning fixes repeated rows, wrong values and gaps. A gap is a missing value. A model learns from what it is given. Wrong data teaches wrong patterns.
A feedback loop uses results to improve the model next time.
Collect and clean the data. Train a model. Improve it with fresh data.
Put each job in its part of the loop.
Loading lab…