QuestionQ77

VerifiedPrepare data and use feature engineering in Snowflake

A Data Scientist is building a new support vector machine model to identify credit card fraud using these features:

  • transactions per_month: the average number of transactions made each month during the previous 6 months
  • spending_per_month: the average dollar amount spent each month during the previous 6 months

The training data contains every credit card purchase with card_id, transaction_id, transaction_dt, and gross_amount.

Which query creates these features so they can contribute equally to the model?

Explanation

Standardizing each per-card feature as (value − mean) / standard deviation gives the spending and transaction-frequency features a comparable centered scale despite their different units and ranges. STDDEV provides the standard-deviation measure used for this normalization.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!