Data Science interview questions and answers

A normal distribution is a symmetric bell curve fully described by mean and standard deviation.

In a continuous uniform distribution every value in an interval is equally likely.

Skewness describes asymmetry: right skew has a long high tail, left skew a long low tail.

A bimodal distribution has two peaks, often because two subgroups are mixed, such as weekday and weekend usage.

A time series is a sequence of observations ordered in time, such as daily active users or hourly revenue.

A trend is a long-run increase or decrease in the level of the series.

Seasonality is a repeating pattern at a fixed period, such as weekly cycles or holiday spikes.

A KPI is a key performance indicator: a metric the business uses to judge success at a chosen grain and window.

Conversion rate is conversions divided by eligible opportunities in a period, such as purchases over unique visitors.

Churn rate is the share of customers who stop using or paying during a period, among those who could have churned.

The central limit theorem says that the sampling distribution of the mean becomes approximately normal as sample size grows, under mild conditions, even if the raw data are not normal.

A confidence interval is a range computed from sample data so that, in repeated sampling, a stated fraction of such intervals would cover the true parameter.

The IQR is the 75th percentile minus the 25th percentile, the spread of the middle half of the data.

Covariance measures joint variability in original units and can be any real number, so its size is hard to compare across metrics.

A z-score subtracts the mean and divides by the standard deviation so values are in standard-deviation units.

Bayes' theorem updates a prior probability of a hypothesis using the likelihood of new evidence, yielding a posterior.

Conditional probability P(A given B) is the chance of A restricted to the world where B occurred, equal to P(A and B) divided by P(B) when P(B) is positive.

Expected value is the probability-weighted average of a random payoff.

The law of large numbers says sample averages converge to the expected value as independent observations accumulate.

Mutually exclusive events cannot occur together, so their intersection probability is zero.