The dataset contains anonymised behavioural data for 11,944 Nedbank customers. It spans up to 34 months of transaction history (December 2012 through October 2015), monthly financial snapshots, and cleaned demographic profiles. Your task is to predict next_3m_txn_count - the total number of bank transactions each customer will make in a future three-month window (November 2015 through January 2016). All historical data is strictly before the prediction window. No target leakage is present in the feature tables.

Data Notes

The prediction window (November through January) spans the South African holidayseason - consider seasonality effects.

TransactionAmount is signed: negative values indicate debits, positive values indicate credits.

This is real-world banking data. As with any production dataset, you should expect missing values, inconsistencies, and fields that require careful inspection. Thoughtful data exploration and cleaning will be rewarded.