Tag: linear-regression
Concepts
Interview Questions
- From five summaries to a forecast, tips and tables
- Rebuild and forecast, factory output from machine hours
- Build and predict with a negative correlation
- Build and predict when the intercept goes negative
- Build the whole regression line from summaries, then predict
- Correlation from R-squared with a negative slope
- Going the other way, R-squared from a correlation
- Correlation from R-squared, a positive relationship
- R-squared from a strong negative correlation
- Recovering the correlation from R-squared and slope sign
- Intercept from averages, a gentle slope
- Intercept when the slope is negative
- Intercept from averages, steep slope small mean
- Intercept with a decimal slope, commute times
- Finding the intercept from the two averages
- R-squared from sums of squares, a strong fit
- R-squared for a weak model, most variation unexplained
- R-squared from the explained sum of squares directly
- When R-squared is zero, a model that explains nothing
- Computing R-squared from the error and total spread
- The line that predicts a forty-year-old is ten feet tall
- When more fertilizer stops helping, a relationship that bends
- The server model that never saw a traffic spike
- Interpolation is safe, extrapolation is not, salary by experience
- Why predicting far outside the data can mislead
- The intercept as fixed cost in a production line
- When the intercept is nonsense, weight versus height
- The intercept as a starting point, seedling growth
- The intercept as a base fare in a taxi model
- What a regression intercept means, the electric bill
- Predicting daily revenue from customer count
- Predicting an exam score from homework completed
- Predicting plant height after ten weeks
- Predicting heating cost with a downward-sloping line
- Predicting from a fitted line, pizza prep time
- A negative residual, the model aimed too high
- A residual of zero, a point right on the line
- A large positive residual in a sales model
- A small negative residual with a decimal slope
- Computing a residual, observed minus predicted
- What an ad-spend slope predicts for new signups
- A negative slope, rent and distance from downtown
- Turning a study-hours slope into a fractional-hour gain
- A depreciation slope, used-car value by age
- Reading a regression slope, cold brew and temperature
- Slope from correlation, sleep and alertness
- A negative slope from a negative correlation
- When the spreads match, slope equals correlation
- Slope from correlation, rainfall and crop yield
- Getting the slope from correlation and spread