From: Inferring psychological traits from spending categories and dynamic consumption patterns
Type | Feature | Description |
---|---|---|
Overall | \(n_{\mathrm{tot}}\) | Number of transactions (log) |
Overall | \(a_{\mathrm{tot}}\) | Total amount of transactions (log) |
Overall | \(a_{\mathrm{avg}}\) | Average amount per transaction (log) |
Overall | cv | Relative spending variability |
Temporal | \(\sigma _{M}\) | Spending amount variability (monthly) |
Temporal | \(\sigma _{P}\) | Spending amount variability (10-days interval) |
Temporal | \(\sigma _{D}\) | Spending amount variability (daily) |
Temporal | \(\mathit{persistence}_{M}\) | Persistence of spending amount (month) |
Temporal | \(\mathit{persistence}_{W}\) | Persistence of spending amount (week) |
Temporal | \(B_{\mathrm{tot}}\) | Transactional burstiness |
Temporal | \(B_{\mathrm{daily}}\) | Daily burstiness |
Category related | \(N_{c}\) | Number of unique categories |
Category related | \(D_{\mathrm{cat}}\) | Diversity of purchased categories |
Category related | \(C_{\mathrm{persistence}}\) | Persistence of purchased categories |
Category related | \(C_{\mathrm{turnover}}^{3}\) | Top 3 category turnover |
Category related | \(C_{\mathrm{turnover}}^{5}\) | Top 5 category turnover |
Category related | \(C_{\mathrm{turnover}}^{\mathrm{all}}\) | Category turnover |
Category profile | \(C_{k}\) | Fraction of expenses (number) in a category k (Table 1) |
Socio-demographic | YOB | Year of birth of the customer |
Socio-demographic | salary | The salary range of the customer (10K intervals) |